Skills you must know before you step into a PhD

I wish I had someone like me in the past to tell me about the material presented in this article. I wish someone could have told me what are the necessary set of tangible skills that I definitely need to know so I can prepare and learn them before I begin to do a PhD.  Furthermore, I wish I had someone like me in the past to tell me about the unwritten esoteric rules of academic life, about PhDs tacit rules, about supervisors & their personality types and about departmental politics and of course about some of the hypocrisy that people might encounter even in the top universities in the UK.

A lot of people are wondering why I am sharing this advice like that freely? Many told me that I should include this article (i.e. the PhD skills article), the literature review article and the PhD assessment and Viva Voce preparation article into a commercial book or at least I should publish them in prestigious educational journals. The advice in these articles is priceless! This is true! The advice compacts my 4 years of PhD experience and in similar vein compacts what I have learned from the experiences of numerous PhD candidates & researchers that I have been in contact with. I have put a lot of advice from supervisors also. Now why am I sharing this advice? I always picture in my mind somebody who is frustrated in a disadvantaged third world country sitting down on his or her PC maybe in an Internet Cafe in a poor village, reading these articles. I want them to be prepared both emotionally and intellectually. These articles are not meant nor dedicated for the spoiled people that have been eating all their lives with golden spoons. The kind of people who have every single advantage that a human being can have!  These articles are purposed for the disadvantaged. They are for pure good karma. All I am asking is a good karma back! Believe me there is nothing here for myself. I have already wasted a lot of time on these articles anyway!

This article is down to earth and talks with guts so if you don’t like that style, well! I am afraid I can not do anything about that. Many lovely academics (staff, supervisors, lecturers) and many PhD candidates from the University of St Andrews, University of Dundee, University of Abertay Dundee, University of Edinburgh, University of Glasgow, University of Aberdeen, University of Manchester among many others, who absolutely loved these articles, have told me that my writing speaks truth without any pathetic sugar-coated language or any form of equivocation and I believe this is what the reader really wants at the end of the day, especially people who want to embark on a very long PhD. People are really tired these days of all the twisted fake or appease-like hedging language that leads nowhere and is pure non-sense and a waste of time. Complete books are written full of non-sense! I have never ever imagined at all that this article and the article on literature review  are now recommended readings suggested by numerous supervisors & lecturers to their current or prospective PhD candidates.

I believe (it is my opinion here) – Yep! I am sorry! I do a lot of annoying introspection & extrospection –  I believe that the fact that these articles are recommended is maybe due to the fact that my articles remind many academics of their difficult time when they were PhD candidates and the suffering, the departmental weird politics and sometimes the hypocrisy, conflict of interests and abuse of power that they had to deal with on a daily basis. I think these articles speak to a part of their academic souls which they liked and have nostalgia for: the part of your soul as an academic that you had when you just started as a researcher: pure intent and “innocence”.

Before you start reading this very long article, Yep it is long! I know! it is worth it! you need to know with certitude that you really want to do a PhD! Actually as a actual litmus test: if you can not read this whole article till the very end and hang in there, you are probably not suitable for a PhD. Why? first I believe that this article contains very useful advice. This is not only my belief or opinion here but the opinions of countless people who read it. Second, you will read a lot of material in a PhD and you have to persevere. If you are not willing to do that, do not bother! A PhD is not for everybody! There is this effect of “monkey imitating behaviour” among the masses where people do a PhD out of jealousy or they think if the neighbour did a PhD,  maybe it is suitable also for them to do! It is not like that at all! I literally heard people telling me that they are doing a PhD just for the sake of having a “Dr.” next to their name on their credit card!!

This does not mean in anyway that a PhD requires a level above the average normal intelligence. Why am I saying this? I can tell you exactly what you want to hear about a PhD: “you will get a lot of transferable skills”,  “You will learn research skills”, “You will improve your prospects of employability and financial security” – The latter one is not necessarily true because academia compared to working in the industry is the poorest and most pathetic choice when it comes to financial security and achieving wealth.  I mean come on! you can get all those skills if you choose another path in Life. What is the real catch in a PhD?  doing research, writing academic papers, presenting your work in conferences to an audience! really! Let me tell you the truth without hypocrisy.  A PhD  is a very very very long undertaking which spans a lot of years (3 to 5 years – full-time) of studying one particular tiny research topic (like eating the same food everyday) and where you should go supposedly  to the edge of the human knowledge in this particular very narrow or limited topic (i.e. at the end of your PhD you should supposedly be THE expert).

Most commonly, a lot of people follow funding opportunities when it comes to research topics so they end up doing a research topic which they are NOT very passionate about. This is quite sad. You must be extremely passionate about your research topic to endure spending a lot of years studying it! It would be very beneficial to you to avoid doing a Ph.D in a topic that you are not passionate about just because you got a scholarship. This might sound understandably very blasphemous to a lot of people. Think of it as eating repugnant food every single day for 3 to 5 years just because someone has offered you to pay you the rent and a small monthly stipend for food! That has been said, a lot of scholarships are not tied to a particular topic and they are usually offered by universities, private charitable organisations or governmental institutions. This type of scholarships is ideal since if you got the money, you hold the power to choose the topic and the right good adviser who is expert in it. This liberates you from working under a supervisor that is not that great or in a research topic that you are not very passionate about.

A PhD candidate a.k.a PhD student will receive a fixed stipend living like a rat for 3 to 5 years compared to being in industry either owning a company or working for one, where she/he will progress on the ladder of life and earn hopefully more money.

A PhD requires a lot of perseverance, diligence and determination and I mean a lot. You might reach stages where you want to give up. In addition, in a PhD you need to manage your time very efficiently since there are no deadlines except maybe your annual PhD reviews, or papers submission deadlines and of course your thesis submission date and  your Viva Voce at the end. So if you are a guy or a gal that needs deadlines and serious pressure to work well, you will find your life to be extremely difficult in a PhD.  The main role of a PhD is to make you an independent researcher and thinker. You have to do all things on your own initiative i.e. being a self-starter, and you have to learn how to be self-reliant. Your PhD supervisor is there only to give you constructive feedback of substance and to suggest research ideas or approaches and to hone by his/her constructive criticism your own suggested ideas based on his/her vast research experience. At the end of the day, you are the one who owns the PhD! It is yours and yours only!

In a PhD, you will meet a lot of very good people, a lot of very smart people, a lot of stupid people and you will definitely meet a lot of hypocrites who mastered saying, writing and behaving in a fallacious and in many cases in an unethical way. The tools of the academic craft of some academics you encounter arered herring’ and straw man fallacies among many others in order for these specimen to survive in academia. I call it a show of PR of knowledge – pompous but empty at the end of the day. It is not just me saying this, Zobel [5] in his book states that straw man fallacy is one of the most commonly encountered fallacy in research and among academics. Eventually, your personal money or scholarship, your most precious years of your adult life… are in the hands and under the whim or mood and ethics of one individual (the supervisor/adviser) who governs the smoothness of your PhD. I will go as far as saying that a PhD in my opinion is a pathetic endeavour to pursue knowledge & innovate, not because it is not noble in what it tries to achieve (i.e in its aims), but because in its current form depends on one person’s mood, human nature, ethics and skills and depends a lot on departmental politics in many cases. This is because you can research, discover amazing things and contribute to humanity’s knowledge as a free spirit outside a PhD. Hopefully you can see my point here! I am not discussing your own competence and your own skills to do a PhD here or the work that is involved in it. You should have those already! Sometimes, it is not about how much hard work you can put in, this does not matter in many situations of hypocrisy or abuse of power and trust. If you are accepted in a PhD especially in a well-respected university that means that you have already all the necessary aptitudes and skills yourself. The system is supposed to be meritocratic.

That has been said, if you are really convinced to do a PhD, please continue reading this article 🙂 Unfortunately, the SEO (Search Engine Optimisation) of this article is not that good yet, so if you land on this article from a search engine, you should get down on your knees and thank whatever deity you believe in. This article increased the traffic on my website of 450%,  so I was forced to upgrade my hosting plan. Probably you landed on a lot of non-sense articles on the web precisely about what is dubbed as ‘transferable skills’ such as ‘leadership skills’ and all this non-sense. It is really funny 🙂 To tell you the truth, those articles are the dumbest material, I have ever read. So according to them, you need to have “leadership skill’ as a skill for a PhD so you can exercise it on your future maybe moody or extremely insecure supervisor. Really! 🙂

You need something more down to earth advice, don’t you think? This article is tangible, down to earth advice. No nonsense! no BS!

NB: I sincerely apologise to the esteemed reader for not having enough time to write a shorter and more compact version of this article. In similar vein, I apologise also for all the typos, grammatical/style errors etc. and maybe weird phrases’ structures; after all this is an informal blog article so I care more about giving you the useful substance more than giving you a Shakespearean language. I hope you will forgive me for those. I assure you that you will not be disappointed of the very valuable advice present here even if it is a bit wordy. Countless people have found this article extremely useful. If you find any mistakes related only to the substance, kindly put them in the comments! Thx

As usual, all my articles, if you have been following my blog, are works in progress: meaning they will be updated continuously. So please always check back for any new material. All the comments are welcomed although I mediate all the comments and do not publish any comment that is just praising me or praising this article or any comment that do not add much. I am getting so much of those and I am sorry that I cannot and will not have whole pages full of comments just praising. Sorry folks! You can never imagine how much I hate praise and hate the people who give praise. Praise corrupts the soul of the person. Please write a comment only if you have any skill, technique or advice to suggest that is directly germane to real/tangible PhD skills so we can all help struggling PhD candidates.

This article is an exposition of very essential general skills that spans a great number of PhDs. The article is more relevant to the PhDs in the natural sciences (Biology, Physics, Chemistry, Astronomy…), formal sciences (Mathematics, Theoretical Computer Science…), applied sciences (Engineering fields, Applied Computer Sciences…) and to big degree to social sciences (sociology, economics, business and management, anthropology, geography, political science, psychology et cetera). Many skills such as mastering LaTeX & BibTeX/BibLaTeX might be considered less relevant to the folks doing research in the humanities. This article does NOT state specific skills but general & very essential ones that every PhD candidate should know.   This article is your first true initiation into what a PhD really is and what you need to know to be prepared technically and emotionally. To be honest with you, ignoring what is written here does not hurt me but hurts you. In either way, my knowledge of the human nature dictates that people need to baptised in fire anyway no matter how much advice you give them in advance. Stubbornness I guess! I also think the ego plays a role here! General and essential in this article means that you still need to learn specific skills and tools needed for your particular research field or for your particular research topic.

Don’t forget that there is a complete section at the end of this article for all the general & yet essential skills that you are strongly advised also to master if you will be doing any Computer Science postgraduate degree and especially if you will be doing a PhD. You are strongly advised to master all the major skills in the main article in addition to the skills elucidated in the Computer Science section so please do not jump ahead directly from this point to the Computer Science section if you are one of ours!!!

In many skills mentioned in this article it is pertinent to have an advanced level in the skill mentioned. The ideal reader for this article could be someone who is (1) thinking of doing a PhD, or (2) it could be someone who was accepted and waiting to start or (3) a first year PhD candidate. It should be mentioned that there are a lot of books out there that cover the topic on how to do a PhD but in my opinion, the majority of them are too general, contain a lot of fluff & non-sense and they don’t give any tangible advice at all. I will suggest at the end of the article four famous books on the topic of surviving the PhD and writing theses that I have read before I started my PhD in case you want to check them.

Please Note: this article is specific to the domain of PhDs/D.Phil (Doctorates of Philosophy in X). There are other doctoral qualifications that are not covered here such a doctorate by publication (a.k.a. by portfolio), practice-based doctorate and professional/industrial doctorate. That has been said, many skills and the majority of advice still apply to the other types of doctorates.

Before moving on with the skills part. It is essential for you to know what are the PhD requirements? How UK universities assess PhD theses? The following PhD requirements section is copied from my other well-received article on PhD assessment and Viva Voce preparations. I have taken material from many UK universities’ policies and procedures on PhD assessment including my university (the University of St Andrews).

Essential Criteria of a PhD

A PhD is a monograph, a corpus or a self-contained piece of work. The following are the basic and essential requirements agreed by a great number of universities on what a PhD should have (sources: in PhD Assessment & Viva Preparation article) – NB:  please always refer back to your own/prospective university policy on Ph.D assessment and read carefully the guidelines that explain the criteria a Ph.D thesis is assessed on.

  1. A PhD MUST contribute to the knowledge of the field of the PhD. Novelty/originality is the central requirement of a PhD. This is the most important criterion that examiners tick on their assessment form. Per example, University of London states concerning the requirements of a PhD: “the thesis shall form a distinct contribution to the knowledge of the subject and afford evidence of originality by the discovery of new facts and/or by the exercise of independent critical power“. Similar requirements can be found in all universities. This does not mean you need to split the atom or have an extraordinary contribution such as a new theory of relativity. There are many ways a PhD can contribute to the knowledge of the topic of research.
  2.  A PhD should have an excellent critical literature analysis that contributes to the understanding of the field, that defends and situates the research gaps and that shows critical engagement with other researchers’ ideas on the topic (i.e. a critical assessment of the relevant literature). This is called the critical engagement with the literature requirement. Some universities refer to this as “situating the PhD research in its general context”  (i.e. in the literature of the field).
  3. The PhD should have a ‘coherent story’, a ‘coherent and consistent narrative’, a ‘coherent corpus’ tackling a specific topic in depth not in breadth. This is called the coherence and consistency requirement. This also applies to a PhD by publication a.k.a PhD by portfolio which involves submitting a set of published peer-reviewed academic papers that follow a coherent narrative which is accompanied by a written introduction and conclusion that pull everything together.
  4. The PhD thesis should be of good academic quality worthy of publication in the best journals of the field of research (i.e. publishable). This is called the quality requirement. This means the thesis has to be well-written. It should be written in an adequate scholarly academic language. The concepts should be well-presented and well-explained and should qualify to satisfy a peer-review. It is extremely essential to publish from your thesis ideally before you submit your thesis for assessment. It helps a lot to even to publish every chapter as one or more peer-reviewed academic paper. This proves in the same time both the originality + the quality requirements  especially when you publish in the best conferences and journals of your field of research.
  5. The corpus of the work submitted should have “the scope of a PhD”. You can find the word “substantial” scattered in the criteria of a PhD in many universities’ assessment policies . This is called the scope requirement and it is the trickiest and most subjective among all the requirements. Usually supervisors/advisors are the ones who are the best to judge concerning whether the candidate have done “enough” already for a PhD. You will probably ask what does “enough” really means? 6 or 7 or 9 chapters? 4 or 5 or 10 published papers? 3 or 5 or 7 contributions? Can we really quantify objectively the degree of the importance of a contribution to the literature? A piece of work could have all the requirements above: originality (i.e. contribution(s) to the literature) + coherence/consistency of the overall argument + critical engagement with literature + publishable quality and still have a minimal scope for a PhD. A high quality Master of Science (MSc) or Master of Philosophy (MPhill) can have all the requirements mentioned above (although not necessarily the contribution to the knowledge requirement) and might not have the scope requirement. Nevertheless, “scope” does not necessarily mean “quantity” since there are a lot of very short and yet very brilliant successful PhD theses of 90 pages or less. You see why it is a bit subjective – it is up to the examiners to decide if what you have submitted has a PhD scope. Usually supervisors should not allow you to submit a thesis that is not of a PhD scope. Scope here means either ‘considerable quantity’ or small quantity that presents ‘research outcomes of substantial significance and impact’. Per instance, in Computer Science, a revolutionary algorithm is worth 10 chapters of small to medium contributions to the literature.
  6. Germane only to UK or similar systems: A PhD should be defensible meaning it needs to pass a Viva Voce. This is called the defensibility requirement which aims (a) to check that the author is the one who actually did the work + (b)  to display the author’s knowledge on the topic + (c) to allow the author to defend successfully (i.e. logically and academically) all the research choices, findings and conclusions. A Viva Voce is not necessary in many educational systems around the world (example: Australia). In some countries such as the Scandinavian countries, Belgium, Netherlands, USA etc. the Viva Voce is public meaning university and non-university members can attend based on some restrictions of course. In the Netherlands, you have to publish from your thesis before you can be awarded a PhD. In UK, a viva voce is a private oral examination. In UK, you do not need necessarily to publish in order to get a PhD. In the UK educational system, a Viva Voce is essential.  This is why on your PhD thesis cover, you would have a phrase such as: “This thesis is submitted in partial fulfillment for the degree of Doctor of Philosophy (PhD)”. The thesis itself, even if it is a perfect thesis, is only a partial fulfillment. The successful defense in the Viva Voce  is the completing piece that allows you to obtain a PhD.

The following is a rendition of few extracts from the assessment reports (reports numbers 2, 3 and 4) given to examiners in the University of St Andrews  to fill out before and after the Viva. The aim here is to elucidate to a prospective PhD candidate (mainly in UK) what criteria a Ph.D thesis and a Viva Voce are assessed upon:

The Ph.D work has to be: (1) lucid, (2) scholarly (3) substantial (4) show original contribution to knowledge. By lucid It is meant that the work is explained very well with an academic language at an appropriate standard.  It is meant by scholarly that the work is well referenced, and of very good scientific quality and style (i.e  it has to be the work of a scholar). It is meant by substantial that there is lots of it or the impact is significant as I said this is very subjective: what does it really mean? so you have to convince the examiners that you have done enough for a PhD scope and that what you have done is an original contribution to the field. You will find terms such as “Substantial” and “original contribution” go hand in hand in many paragraphs in these reports like “substantial contribution to knowledge” i.e. a lot of contribution to the field of research.

The following quoted phrases are taken from the reports that the examiners complete in the University of St Andrews before a Viva and at the end of a Viva (the purpose here is just to elucidate to you what you need to be aware of):

  1. General quality and originality of the research described in the thesis
  2. state what are the main contributions to the field that are made in the thesis
  3. where relevant how the quality of these contributions compared to those made by other PhDs in the subject you may have seen or examined”- This is an interesting one! Not only the thesis has to be of a good quality but of a quality that is better or at least similar to all the PhD theses that the examiners have examined in the past.
  4. What if any of the thesis work was published or is of publishable quality – this is why having publications from your thesis helps a lot in defending successfully a Ph.D thesis. If you publish, examiners would tick this and write down what you have published.
  5. whether English has been to an appropriate standard” – this is why you need to have a very good command of the English language precisely the correct academic language. Mastery of academic language is learned through time by reading books that teach it. I will suggest some later in this massive article. In addition it is gained by reading tons of academic papers and by learning from the language used by many scholars in your field. The more you read, the better you become.
  6. any concern that is raised in the oral examination has been addressed” – Recall the defensibility requirement. This is why we have a Viva Voce in UK so here the examiners write comments on how you defended certain raised issues.…

In report 4 which is the joint report compiled by all the examiners after Viva  and which is given to you the PhD candidate (in the case of the University of St Andrews). The examiners have to discuss how the oral examination or Viva Voce went between each others and answer questions such as:

  1. Does the thesis show evidence of originality in the work described?
  2. Is the literary style and presentation of the thesis satisfactory?
  3. Does the candidate possess an adequate knowledge of the field of study and related literature?
  4. Were any concerns highlighted in the individual examiners’ reports satisfactorily addressed in the oral examination?
  5. How were any significant differences between the examiners’ reports resolved?
  6.  “Does the abstract appropriately reflect the content of the thesis?

WARNING: I should mention here a common misconception that some PhD candidates have. Some think that if they publish many peer-reviewed academic papers that they would have by consequence some kind of a very secure position and thus they become arrogant and feel that the Viva Voce is kind a promenade in the park. I know a person in an English university who published many papers in quality journals and still got major corrections after her Viva. Publishing from your thesis just tells the examiners that other peers (i.e scholars) have looked favourably on your work nothing more than that. There is a lot of important criteria of PhD assessment other than “originality” that your work might/or might not have. It helps enormously to see that you have published many peer-reviewed papers from your Ph.D research since it shows there is original quality work. Nevertheless, always approach the Viva Voce with the utmost humility possible and maybe also with a bit of healthy dose of fear and know that no matter what the whole world or your supervisors think of your work, it is up to the examiners to give you a Ph.D or deny it from you. In other words, you have to convince them and defend successfully your research choices\methodologies\conclusions.

Now with the skills that are compiled from years of PhD work!!!!

The following are SKILLS not specific to a particular domain or field that every PhD prospective candiate is strongly advised to have:

Skills you MUST know if you plan on doing a PhD

DO NOT STEP INTO a PHD in a reputable university without knowing the following (Warning: mastering the following skills do not apply to “peanut butter” research in low quality universities):

PhD General Skills
PhD General Skills to know/do before the PhD

A – Master knowing how to choose the right and good supervisor/adviser & how to choose a good PhD topic

Important Notes before we start: This section covers one of the most important skills if not the most important skill of a PhD. The one which if discarded, leads to severe repercussions for the success of any PhD. How to choose a supervisor?! What type of supervisor to choose? What is covered here pertains to traditional UK PhDs where in such PhDs, you are the one who choose at least your primary supervisor. You are the one who usually approach an academic staff about a PhD topic. You both end up working on a PhD research proposal. The information here might be less relevant to industrial doctorates or other type of doctorates where your options are limited regarding the choice of a supervisor. In the UK system, there are two types of supervisors: primary and secondary. In the majority of cases, the primary supervisor is the most important in terms of control over the PhD supervision and the secondary supervisor role is more of a pastoral nature.  That has been said, I know in some universities and schools both could share supervision with a certain assigned percentage (60/40 or 55/45) or even equally. A friend of mine told me that during his PhD, the secondary supervisor was more important than the primary supervisor and this is because the secondary supervisor was a previous PhD supervisor of his primary supervisor so the old dynamics of power were kept. From my experience and the experience of numerous PhD candidates: it is always advisable to choose at least one supervisor who is very technical and could help you on the ground level of your research topic. In case you will have a primary supervisor who will be giving you  high-level advice, it is advisable to have a secondary supervisor who do the opposite & vice versa. Supervisors who give high-level advice does not mean giving you only language advice or advice that lack research substance. You should not have such supervisors because they are literally useless and even harmful – I will talk about this later in this section.

The relationship in research degrees between a “supervisor” (a term used in some countries such as the UK) or an “adviser” (the US term) and  a PhD candidate is more like an apprenticeship relationship than anything else meaning a relationship between an experienced research mentor” and a research “protogé” or “mentee”. When embarking on a research journey, one would need an experienced individual to facilitate solving hurdles, to guide and to support. Unfortunately, universities are now more driven by money these days and they lost their initial purpose. This also affected the role of advisers which became more diluted and superseded by different roles such as what some supervisors consider themselves as  being more of “managers” with sticks and carrots not mere “mentors” (as it should be and used to be). A PhD adviser is a person that in theory  you can easily change the same way you can change your shirt especially if that person is not useful anymore or is wasting your valuable research time, and your funding money (usually the money of the tax payers), or more importantly when he/she starts to behave unethically which is not uncommon.

According to the UK Quality Assurance Agency for Higher Education doctorate guide [2]  your supervisors should act as “expert guides and professional mentors as you develop both your research project and yourself as a research practitioner”.

Since you are working with a supervisor for 3 to 5 years (full-time scenario), you need always to aim for the [Good  AND (Boolean Operator) the Smart adviser]. Supervisors are divided – unsurprisingly! 🙂  into four categories: [Stupid and Good][Stupid and Bad] [Smart but extremely Egomaniac and Evil] and [Good and Smart].

[Stupid and Good]  supervisors do not help you much in anything really, they might facilitate only bureaucratic processes and [Evil and Smart] is a big big big trouble, the latter category tends to have supervisors that are extremely exploitative of their students and very grumpy, moody, insecure and childish. Actually all the sets above in which you have the variable [Stupid] in them are detrimental to you and normally this category contains a lot of lone wolves (academics who are not part of any research group or do not have a research group under their supervision). That has been said, there are a lot of lone wolves who are [Good and Smart] but the general tendency of this category is to have the Stupid Variable in my experience :-).

[Stupid Advisers/Supervisors] usually look at you with those stupid eyes and stupid faces, insignificant is their contribution to your learning and knowledge. Insignificant is their technical advise! to the point you ask yourself: hmmm..! what would be the difference, if I had my grandma as my PhD supervisor! Would anything change? Sometimes the only advise and feedback you get from these specimens is in the English language such as whether your spelling is right or wrong and where to put articles such as “the” or “a”. I am not kidding!. It is not their job to look at the language of your reports or academic papers. This is not the kind of feedback you are waiting to get from an “experienced research mentor“. Their job is to give you constructive feedback of substance on your work since this is why supposedly they are experienced researchers in the field in the first place. It does not harm if they do give you language feedback in addition to the more important useful feedback of substance especially in the writing up phase. These types of supervisors have what is dubbed in Psychology as Dunning–Kruger effect. Bare in mind that at a certain late stage you will become more knowledgeable in your specific research topic than your supervisor but you always need him/her to give you constructive feedback of substance. Feedback of the adviser is very essential in all the years of a PhD especially at the end but the type of the feedback at the end of the PhD is usually more concerned with how to defend effectively your research in your thesis and your Viva Voce and how to make sure that all your conclusions are solid and well-founded and to make sure that your arguments are clear and cogent.

In all UK Universities, there are normally complete institutions that help students for free with academic language. Per instance, in the University of St Andrews, there is the lovely  English Language Teaching  (ELT) department.

The truth of the matter is that a lot of such supervisors normally piggyback on other supervisors’ backs in the department whom are usually more intelligent. Some also piggyback on the work and intelligence of their PhD candidates or postdocs. So you might see that they have an “impressive”  list of publications but this is all fake. This is in order for [Stupid Supervisors/Advisers] to survive academically since we have the motto: “Publish or Perish“. They use the technique of: “Quid pro Quo” or “Scratch my back so I can scratch yours” especially in publications. They are the most pathetic specimens you can ever encounter and I wish you never encounter people like that. This is the type of UK supervisors/advisers that when things does not work as they should, they would throw the matter on you usually ignoring their incompetence or lack of skills in supervision and even in the research field itself.

[Smart Supervisors/Advisers] on the other hand, always always always know how to make things work in a PhD. They know what actually works in the research field and what is worthy spending time on investigating  and what does not work and is a waste of time and money. Of course you have to do your own PhD, no one will do it for you, but they are the people that would spot a barren research path quite quickly and intelligently and this is because they are extremely experienced. When you sit with them and you both discuss ideas and matters, you have your head full exploding with ideas in contrast when you sit with a moron, you have normally nothing. A hour with a [Smart Supervisor/Adviser] is worth 2 years of meetings with a [Stupid Adviser/Supervisor]. If you are a good, clever and hardworking PhD candidate, you will always always always get a PhD with a Smart Supervisor no matter what unless something out of your control gets in your way.  Not to mention, [Smart Supervisors/Advisers] open usually a lot of doors for you in terms of helping you publish in conferences and journals, in providing you with connections (very important), normally they possess and procure many expensive equipment and resources that you can take advantage of.  You should think of all of what I am saying here when it comes to choosing a supervisor.

A scholarship and a monthly stipend is the money dedicated to put the food on your table not the money that should be spent on anything else. It should NOT be spent on buying books or equipment or software licenses or Laptops or PCs or servers. You need an adviser/supervisor that provides you with all the resources and equipment needed from funded research projects and who facilitates also the access to people’s skills usually via a research group of PhDs, Masters and Postdocs or connections to the industry especially for PhDs that requires such access (i.e. scientific PhDs). Departments are usually hypocrite, they normally give you at the start of the PhD, a rotten old PC used by someone else and a desk and say hey! you are settled now! No you are not!

One thing to put always in mind, that if you have a scholarship even if you have obtained it with the blessings of an academic. That scholarship is NOT tied to that academic blessings and is NEVER GIVEN TO YOU BECAUSE OF HIM OR HER at least in well respected universities. It is given to you because you deserved it academically. It is a meritocratic system!!! I am saying that because a lot of people that I know, who come from eastern cultures both Far East and Middle East, feel somehow they are indebted to the supervisor and thus they remain with such supervisor even if they are not reaching anywhere or even if the supervisor-candidate relationship is very unhealthy. Move on!, a scholarship is the same as your own money. Normally this money comes either from the EU, or from Companies, from charitable organisations, from foreign governments including the candidate’s own government or from UK research councils meaning UK tax payers. So please take care of where you spend it and with whom you spend it with. Don’t spend it with a moron supervisor. It is unfair to do otherwise, it is a hard earned money and the funder and the university are investing in you NOT in your supervisor/adviser. So please be at that level of responsibility!

A supervisor/adviser is nothing but a human resource dedicated to help you. It is after all your PhD, your money, and your time taken from the most important years of your life. When people come to me and tell me about the problems that they are having with their supervisors whom either have sever egomania, control freakishness or psychological insecurities or they do not give them any descent feedback of substance or who ignore their students for a long period of time. I usual told the candidate to fire his/her supervisor (figuratively speaking) and to quickly find another suitable and more balanced experienced academic. At the end of the day, it is your PhD not theirs and you will definitely be blamed and you will be judged and examined in a Viva Voce at the end not them.

As I said previously always choose a supervisor/adviser who has considerable research money. There is a big probability that this means he/she is smart. Not necessarily though: if you live long enough, you see too much. This type of supervisors have complete research groups with many PhD candidates and maybe also many postdocs. This means that you will have access to a lot of resources, machines, equipment, expensive software etc… and more importantly access to other people skills and expertise. Also this opens opportunities for you to publish papers or to contribute to other people’s work in the form of collaboration. Pay attention in this category you have the other parameter to consider: Bad vs Good. If your luck happens to be terrible, and you lend on the [Bad] in this category, your life would be literally hell especially if it happens your luck was with those dubbed as the ‘untouchables supervisors‘.

Definition of untouchable academics: they are academics with severe vitamin D deficiency for lack of exposure to the Sun or lack of interaction with nature and its harmony ;-), who are usually psychologically insecure with unfulfilling lives  and more importantly who bring yearly millions of pounds worth of research money to your school to the point they become ‘untouchable gods‘ (or at least this is what they think). A lot of them have as I said unresolved psychological complexes. The school has to keep them happy because of funding cash flow. So be careful with such a choice, always look for the [Good] trait, because the [Bad] is literally lethal to your PhD. In addition, always aim for the supervisor who has an impeccable record of PhDs completions i.e. students tend to finish successfully their PhD with them and more importantly successful completions in the funded normal duration (this is super super important!).

You can change your supervisor of course during the PhD but this is usually involves a big headache and you should have a legitimate and convincing reason. It can be difficult in some universities or departments. You should know that even in the worst case scenario: where no supervisor can be found in the department with the right expertise, the school can assign you a supervisor from outside the university.  So the impression that there are no advisers other than a certain person in your department is false. How would be picking later your Viva Voce external examiners in your field? just have one name less on that list of prospective names that you might consider at the end of your PhD  and have him/her as a adviser. Every problem has a solution, the most important thing is to work hard, move forward and get the PhD! The school, equipment, funding, the adviser etc… are only resources to help you. Always remember this and don’t be convinced by anybody that it is something else! Pay attention: Some sicko supervisors like to sell you this idea that they have control over you, that they can destroy your PhD! they don’t!

I did not yet discuss two types of supervisors that you might encounter in a PhD:

1) Territorial supervisors or supervisors with “control freakishness” syndrome who want to control the whole supervision and to control all the research ideas and would literally hate you when you take any advice from any other academics even sometimes from your own second supervisors. You need every opinion and every advice you can get to be able to have the constructive & healthy criticism that you need  to defend your research in conferences and more importantly in your Viva Voce. The primary supervisor opinion or his love or hate to you  is inconsequential and sometimes useless when it comes to your defense in your Viva Voce. I have been told about this phenomenon by some supervisors from the University of St Andrews and from another Scottish university who were secondary supervisors to students. This phenomenon is know as “being a control freak” or “being territorial” . Avoid choosing such supervisors because they are usually dangerous and very unstable and in few concrete cases that I have witnessed they were found to be extremely unethical.

2) Another type of UK supervisors that you might encounter is the one who is very sneaky and very corrupt who – believe it or not-   don’t want you  to finish your PhD exactly on  time even they would claim the opposite to you. It is NOT helpful to them and it does NOT serve their personal devious interests that you do finish your PhD on time. They make you spend more time working on side research projects that benefits them or their research agenda and do not necessarily benefits you or lead you to a successful Viva Voce! The official terms here are masked exploitation and abuse of trust. This was found true from the experience of some PhD candidates in official complaints. This is when you trust the supervisor’s  judgement in the PhD and it turns out he/she has other agendas that are NOT  in your interests. We are talking here of course about normal PhDs and not necessarily about industrial ones. This type of supervisors considers PhD students as “cheap labour that are very useful in doing their own projects. PhD students are usually more desperate, more docile & more servile than postdocs or other types of employees so they can be taken advantage of very easily especially international students. An international student have an additional fear which is that his/her Visa being curtailed if the supervisor pulls out the plug on PhD supervision. In addition, PhD students trust by default blindly their supervisors. These types of supervisors keep adding supposedly “new research requirements under the guise of “this will strengthen more the PhD even when you have clearly reached the “enough contributions stage“.  Have a look again at the start of this article to learn about the requirements of any PhD or consult my well-received article covering thesis assessment & Viva Voce preparations. The projects requested by such supervisors may not even relate at all to the research topic. This actually happened in my own university! They can also be projects that are divergent research-wise  or in the best cases they can be tangential to the main research topic. So please please pay attention!!! This is a form of exploitation and abuse of trust disguised under normal PhD research. Always consult other academics in your department and please consult your secondary supervisor(s) about everything you do or being asked to do by you primary supervisor and please don’t accept to work on any projects that are either tangential or not related to your main topic of research. It burns your precious time and funding money and will lead you to failure if not rectified quickly and swiftly!

To avoid all the headache of changing a supervisor in the middle of a PhD, take your time in choosing a good and smart adviser from the start of the PhD!

Few horrible examples from experiences of friends and colleagues so that you are aware of what could happen to you. These experiences are authentic and happened in “high quality & high ranking UK universities”. I am stating all the incidents with the permission of the victims.

A PhD candidate in a UK university (one of the top 3 in the ranking), tried to explicitly resign in an explicit audio recorded meeting from a company owned by his PhD supervisor. He happens to be a PhD candidate doing a PhD that is NOT industrial and he happens to work in a software company owned by his supervisor at the same time.  He was audio recording all his PhD formal supervisory meetings. Before I forget: do not work in a company owned by your supervisor  even if you are tempted  (assuming you are not doing an industrial PhD) – there will be definitely sooner or later a clear clash of interests. Not to mention this is will complicate matters if you would have to submit later an official stage 2 complaint to your university (since you would be considered a staff of a company (i.e. separate work policies that could be outside the University control) and a Ph.D candidate (internal policies of the university) in the same time. In addition, employees have tribunal courts to take cases to. University policies should be clear on cases of clash of interest where PhD students work in supervisors’ companies. It is bound in such cases that the work for the PhD is mixed up with the commercial work that benefits the company and supervisor’s pockets.

Back to the story: this PhD candidate wanted to resign and stop working in the company so that he can focus solely in the remaining time and with the remaining funding money left on his PhD. His job in the company was to create silly websites for the company, a work that is not related at all to his PhD. The supervisor after hearing the resignation (i.e. the resignation was on the audio recording), have the audacity to force him to turn off the recording and then blackmailed him and asked for a quid pro quo: He told him “Do your own bargain and I do my own bargain”, “I have a commitment to you so you have a commitment to me“. In other words, he is saying to the candidate “stay and work on crappy projects for my company that are unrelated to your PhD while you run out of time and money in your PhD and I will keep supervising you“. Well! the relationship is not a “quid pro quo” or a “scratch my back so I can scratch yours“.  The supervisor “bargain” is his job, his duties as a supervisor/adviser of a PhD, he do not need a noble price for it and he MUST do it, whether in the case of this PhD candidate, he do not have to keep his “bargain” especially if his PhD is in jeopardy (running out of time and money). The candidate did not even agree to such bargain which is a sick and weird term to use anyway. The candidate does not have to work on projects that are not related to his PhD. It is like dealing with a thug drug dealer. Supervisors who blackmail their students are academic thugs this is what they are!!! 

This PhD candidate  used to tell me that whenever this cheap supervisor used to be hungry, he came to this PhD candidate desk and tell him that he is hungry explicitly, keep going back and forth with the same message and because this PhD candidate is from a very hospitable and proud middle eastern culture, he used to buy his supervisor hundreds of Pizzas throughout 2 years. The amount of money spent just on pizzas is a lot more than the monthly salary that he used to earn from this crappy company and the peanut butter silly projects that he was working on in the company. Not to mention the gifts such as bottles of wine this  man used to accept, the PhD candidate was trying to appease him and this is the problem actually, – you do not appease such behaviour, you put an end to it soon! Not to mention, the extremely abusive language and behaviour that was used against my friend in terms of continuous use of the F word regularly in supposedly professional formal supervisory meetings in this UK university, words such as “fuckhead“, “fucking“, “you are fucking agitated“,  even the supervisor farting in one meeting without excuse  raising a little bit his “arse” from the chair (I will let you imagine that!) and without going to the toilet and telling the PhD candidate that “your supervisor is talking from the arse” . The supervisor told his PhD candidate that he is his slave. There are many other examples that could not be said here as if what I just said is not hair-raising enough for the reader!!!.  Many of such behaviour were captured on audio recordings, other behaviour was not audio recorded but was clearly documented in emails.

This cheap specimen , this cheap supervisor used to come and put his legs on this PhD candidate’s desk facing toward the student. The latter boiling from inside of course, was under blackmail and fear of dropping his supervisor so he does not lose the PhD. This PhD candidate witnessed  even more horrible stories but for brevity I will not include them here. This same specimen of supervisor asked him to do a work for another PhD student that was even published before without telling him explicitly, nor implicitly that it is actually is for another PhD student nor more importantly that it was published before.  I will remind you again that one of the major requirements of a PhD is originality (i.e original contribution to the knowledge of the field). So this supervisor can not force his PhD candidate to do a work that was published by another PhD student. Where is the credit? The university is the one with very low quality standards for hiring such a man to deal with students in the first place. There is no academic ethics at all.

This specimen supervisor – As you can see, I am intentionally avoiding using harsher adjectives here as much as I am tempted to but I ask the readers to choose the right adjectives for this supervisor’s low quality behaviour based on their moral compass and decency –  This supervisor is known by the majority of people in the institution as an extremely cheap and a low quality individual but is kept because unfortunately he and other similar academic specimens have long UK work contracts with weird laws and are rotting in the different schools of this supposedly high-ranked  high-quality university.

This  married unprofessional man  allegedly (actually highly probable) has a pathetic “school boy” crush,  and probably more than that, on his female PhD student, the one that he is asking another student to do her work. The supervisor in question is writing her academic papers so that she can get undeserved funding and even asking his employees in his company and as I said his PhD students in a veiled and sneaky way, to do work that she would end up claiming credit for in a way or in another. Actually this girl explicitly told the wronged PhD candidate that the supervisor wrote her academic paper, the one that she submitted with her PhD application to gain acceptance and funding. As usual these things (i.e unorthodox relationships) are difficult to be proven with evidence. I mean it is difficult to prove a love/crush kind of relationship such as that but one could easily prove  clear cases of favouritism resulting from relationships of this kind. There should be no favouritism of a PhD candidate over another of the same adviser under any circumstance and for whatever reason. Nothing wrong with love between two people (actually it is a beautiful thing) but this should NOT be at work, there are clear rules that govern that.

Always bare in mind that favouritism in all its forms and types is against most UK universities’ policies and even if your policy does not cover that it is against UK law. You might see favouritism  when a school spoils one PhD candidate with a lot of equipment and money while giving scraps to another (it is not uncommon at all). Any dating or unorthodox relationship (even with consent) between a supervisor/staff and the student if proven is a fire-able offence for both.  Supervisors/Advisers can NOT date their students without following specific procedures,  without ending the supervision relationship and without notifying HR. 

Another horrible example that was experienced by another friend of mine, I can only say that she is a lovely gal from New Zealand so that she can know that I talking about her here with her permission: her supervisor (a clearly pervert and deviant man) sexually assaulted her many times, initially he used to look always at her “boobs” -sorry for the expression – in formal supervisory meetings not her face like he is some kind of teenage pervert school boy. He made different approaches that she rejected vehemently on many occasions (until she complained about it explicitly) also he forced her to put the name of some academics in her department on her academic papers (they were like a thug cartel). This is called abuse of power [5] where the names of senior academics are put on papers that they did not contribute to and this is very common. Conflict of interest, harassment and abuse of power are inter alia major issues of ethical concerns in academia [5]. These thugs have contributed  nothing to her work just so that he can get free academic papers back from them and their students. How pathetic is that!!!  This happened in United Kingdom (not a third world country)!,  precisely in a very well-respected Scottish University. She complained and he was fired after a long tedious investigation process. The university has a zero-tolerance policy against abuse, blackmail, exploitation, favouritism, and harassment of all kind especially the sexual kind. I respect her and respect her strength and courage for coming forward and thank goodness that scum got what he deserves. You can never begin to imagine the number of women who are victims yearly of such abuse and harassment in companies, public institutions and universities whom do not dare to do anything. It is really horrible!

Some universities unfortunately in the worst horrible offences (we are talking UK and US universities here), exercise only minor sanctions that affect the progress of the thug academic on the organisational ladder as was witnessed in some cases and this is a “not enough” behaviour from the part of supposedly high-quality reputable universities since a lot of cases merit and warrant strongly and clearly that the academic to be fired otherwise it will embolden other exploitative abusers, blackmailers and harassers.

I have been told by some academics that the culture of academia in UK is a culture of reputation so firing an academic does not destroy them. Acts they have committed destroys their reputation in their departments and throughout the university and other universities and among funders even if they stay in their jobs as shells. No one will hire them anywhere else. That is the real death sentence! It is more convenient that the reputation of such scum bag academics to be destroyed so that they reach a stage where they wish they were fired in the first place. They deserve to end up begging on the streets.

If you have been wronged by a UK university staff member (including supervisors), a lawsuit might be suitable option but follow first the normal path: complain to the direct manager (a.k.a level 1) which in the case of supervisor, the head of school then if not resolved you can file an official complaint at stage 2 (a.k.a Level 2) which will include an official investigation. Do not forget to submit as much evidence as humanly possible (emails, audio recording etc.) If you are not happy with the resolution or there is evidence of a cover-up to protect the academic or the investigation has committed logical fallacies – one common fallacy is the straw man fallacy which involves either misrepresenting the victim’s arguments or the evidence or to dilute the evidence and then attack the diluted versions or to focus only on one component of the complaint with has the least concrete evidence while disregarding the more serious components with stronger evidence. Many complaints are buried because of not enough evidence or lack of tangible evidence. This is why I always urge prospective and current PhD students to record all their formal supervisory meetings and to document things by emails. Your evidence should be legally submissible. You never know the circumstances that might oblige you to go to a court.

If you are dissatisfied with the outcome of your stage 2 complaint, you can  go to the governmental overseers of the Universities (their bosses if you want). In Scotland, there is the Scottish Public Services Ombudsman (SPSO). In England and Wales, there is the Office of the Independent Adjudicator (OIA). SPSO and OIA will investigate your case if you are not happy with the university’s outcome. The university has to give you an official written  outcome at the end of the investigation containing usually one of the following terms: Upheld (or Justified), Partially Upheld (or Partially Justified), Not Upheld (or Not Justified).  Make sure that the university will do that from the start of the complaint because OIA and SPSO needs to see a clear written outcome in order for their processes to start. In one case I heard of, the university intentionally did not give a clear outcome and kept blabbing to the student about how it can not do that under DPA (Data Protection Act) legislation. They knew that their staff member is extremely guilty and thus halted the process for many months so it did not go to OIA keeping their student in a horrible emotional condition reaching a breakdown. Imagine the kind of sneaky behaviour from some universities! Eventually it went to OIA and they were forced to pay a big amount of money (5 zeros number) to the student as a compensation for all distress, extreme disrespect and inconvenience caused by such unprofessional behaviour and the university was forced to fire the staff member after a very long process. This university is among the top 10 universities in UK so I am not bringing an example of universities which resides low on the ranking ladder.

Check this article: OIA forced universities in many cases to pay  compensation reaching up to 400000£.  If you go along the path of OIA (England, Wales) or SPSO (Scotland), it is already so embarrassing  for the university  anyway since they aim usually to keep things without damaging their reputation in other words they aim to keep “their house in order“.

If SPSO or OIA folks per example, discovered any cover-up that the university is involved in (it has happened a lot before!), or they discovered any disregard of any shred of evidence, any discrimination, any incompetence in the investigation process and or in the investigative team (like assigning the wrong guy or gal with the wrong expertise to be investigators,  bias  etc..) they are in lot of  trouble. Trust me! and their reputation will plummet to the bottom not to mention that this will give you a strong ground for a lawsuit that even SPSO/OIA folks will help you with.  If by that time the thug academic is not fired, the judge will force a termination of contract in addition to whatever the judge feels necessary in terms of damages and compensations for the suffering that this has caused you emotionally, and the loss of big amount of time and money.  As if this is not enough, usually the media will pick up on the story very quickly  especially in a supposedly highly-reputable university with supposedly high moral compass and thus another death sentence! Mishandled genuine grievances can lead to massive scandals that end academic careers. Justice is always served no matter the amount of time it takes for Lady Justice to exert her power and no matter how much conspirators try to stand against her!

I think there is a good reason to mention the above cases that  I have been told about especially to international PhD students. Why? I am proudly from an eastern culture so I can say this freely. A lot of international PhD students mainly students from eastern cultures have a weird sick and unhealthy dose of deference to professors or doctors or elders, sometimes this go to the level of being servile or docile to them instead of being critical and skeptical people who should refuse requests that are detrimental to their research or requests that are out of the decorum. Easter people do not know how to say NO. I heard a case of a PhD candidate who s doing the laundry for her supervisor in a well-known Scottish university 🙂 Come on, I mean this is extremely sick! I do not know if it is just hearsay funny gossip or she is a friend with her supervisor and even friends do not do laundry for each others. International students mainly of Eastern background think that if they argue with their supervisors that this is considered a rude gesture. A lot of students even avoid asking questions. This is due to the fact that generally in Eastern cultures, the elderly or the knowledgeable/wise is very revered almost deified.  This should not be the case at all, you have to respectfully refuse or respectfully draw your red or even yellow lines in the sand and be very critical of everything including your supervisor’s character or behaviour.  I am using the word “respectfully” here to emphasise that you should always respect people to receive respect back and also I do not want you to smash the face of your supervisor if he makes a mistake even if he deserves it 🙂  Trust me there are many legal and policy-wise ways to punish bad behaviour coming from supervisors or staff.

Many filed misconduct cases show that this “weird behaviour of deference” from international students is easily exploited or taken advantage of, especially from bad, cheap and low quality UK supervisors. As an international Ph.D candidate, you do not have to have any unhealthy weird deference to anybody even to the principal of the University! Call the principal by her first name when you send her an email or see her!  The Ph.D aims to create out of the candidate a critical independent thinker. It aims to make out of you a skeptical intellectual i.e. an individual with a unique voice. Check out the University of Glasgow video touching a little bit on the subject of “deference aspect” that differentiates international students from UK students.

How to approach the matter of choosing a supervisor intelligently

What you are looking for here is two important sides of the prospective supervisor/adviser: the human/personality side and the technical expertise side. Both are equally important.

For the technical expertise side, it is quite easy to figure out if someone is suitable. You should read all his/her academic papers, all books, books’ chapters or at least read the most recent published work or most relevant work that you are interested in. Some academics have thousands of papers so it is not reasonable to spent months reading a prospective academic’s papers and at the end, you discover that he/she might not be interested in supervising you.

PS: Please bear in mind that having a big number of published papers is NOT necessary an indication of intelligence nor of competence.

Some disciplines or fields have too much fluff and padding in their literature 🙂 while others don’t have that. In addition, an academic in medicine might have only 6 or 7 high quality peer-reviewed published papers and is considered an authority in his/her field while another with 500 papers is not considered an authority. Number of papers do not mean anything! It is the quality of the academic papers, the quality of the journals where they are published in and more importantly than all of this is their impact that matters. Not to mention that you see a lot of Piggybacking in academia or “scratch my back so I can scratch yours”  so a lot of academics end up on academic papers which they have no clues what they contain. So be aware! Check always papers where the “prospective supervisor is the first author”.

I advise you also to check the undergraduate/postgraduate courses, tutorials, labs, workshops that the prospective adviser/supervisor have taught in the past or is teaching now, and check also the presentations, the talks, the grants etc…

You might not have access to a lot of published papers since you could be outside an academic institution. A lot of prospective supervisors are happy to send you a list of the papers that you are interested in. Sometimes they would send you the final draft manuscripts so that they understandably avoid headaches with publishers. Many upload manuscripts to ResearchGate or You can always ask a friend who has access to download the papers and send them to you. I am not expecting prospective PhD students in poor third world countries to buy 5000 pounds worth of academic papers so they can write their research proposals. Actually I am not expecting anybody outside academia even from a wealthy country to be able to do that. Expecting that is extremely hypocritical and ridiculous!

For the human/personality side: You should always aim to choose a good and well-balanced human being as a supervisor/adviser so to avoid all the headache you just read about in the previous paragraphs and please don’t say that these cases aforementioned might not happen to you. They happened to a lot of intelligent, good and hard working students in highly reputable UK universities.

Now if you are outside academia, you would never know whether a supervisor is a descent balanced human being from just looking at his/her profile web page. Can you be sure that I am not a serial killer!? How would you be so sure! Well, let me clarify this fast, I am not! 🙂  The point I am driving home here, is that this type of knowledge is only gained on the ground and so your aim before investing 4 years+ of  the most expensive years of your life and before investing your money (whether self-funded or through a scholarship), is to seek advice from the people who have that kind of knowledge. By the way, a lot of idiots do not know that if you have a scholarship no matter who you are doing your PhD with, it is still your own money. Funders are investing in you a lot of money not in the supervisor/adviser, they do not care if you do your PhD with Joe or Micheal or Monica. For them, they need the job to be done. So a scholarship is your own money that you deserved it not that of a supervisor/adviser (the only exception is if the money come directly from the supervisor own salary which is impossible).

(1) The first advice I can give is that you must ask academic staff with whom you have good relationship with, about the person you have in mind for a PhD supervision. What are the pros and cons of that person? The cons are more important that the fluff and the pros. Bare in mind, a lot of people are very very cautious to tell you negative things which is what you are after. Try to show them that you are trustworthy, that you are not talkative and that you are only asking so that you do not regret later choosing this supervisor later. They will understand and start telling you the real truth about the person.

(2) Ask also PhD candidates about their advisers/supervisors but DO NOT ASK  FIRST or SECOND YEAR  PhD students (they are kissers of hands of their supervisors – so I do not use another word here). Ask LAST year PhD candidates who are just near submission or better, the people who have finished their PhDs with the adviser in question or even A LOT BETTER, people who have switched to go with other advisers/supervisors because of a problem since these people do not care anymore when they talk about the negative aspects of their previous supervisors. Finally, ask normally someone like me who does not give a damn about the world! and about the hypocrisy in the world!. Find someone like me in the school of your target.  It is impossible that you can find a person in the history of humanity, who do not have negative traits. I hate when I ask first/second PhD students about their supervisors and the majority of them pathetically could not give me a single negative thing. This is unnatural, stupid, pathetic and hypocrite. Every person has flaws PERIOD. Every person has a negative aspect of their personality.

(3) Do your own calculations of risks & ask around:

  • (A) How many students have successfully finished their PhDs with the adviser in question?,
  • (B) How many have dropped or not finished and Why? It is very important to know the Why?  Knowing why students have switched to other supervisors allows you to detect the nature of the conflicts and the personality of the supervisor.
  • (C) What was on average the duration of the PhD of this particular adviser? If long duration on average but with success at the end, it means the supervisor/adviser is either [EVIL and SMART] or his/her topics are just bananas. Long duration is not necessary a bad indication. After all, it is research! Nevertheless, if all PhDs takes more time that what they should take, this means trouble ahead! This should be alarming to you. It is a probable indication of Evil because probably supervisor is forcing his/her students to work on matters not related to their research.
  • (D) Look at the ethnic, gender and religious composition of his/her PhD students: If no Arabs, no Muslims, no people of different skin colour, no Jewish and the supervisor/adviser has been supervising for a very long time, that means there is a strong probability of having either racism, or Anti- Muslims/Arab bias or Antisemitism. If there are no women between the students, and he is a man, that probably means that there might be a sexism issue or he is a religious fanatic!
  • (E) Does the supervisor give his/her PhD candidates enough time and effort in terms of guidance and supervision? Does he/she ignore them for a long period of time? Does he/she give them useful feedback of substance?

Bare in mind that in many  universities’ policies, there is a minimum number of formal supervisory meetings per year for full-time and part-time PhDs. Anything less than the minimum means that you & your supervisor are breaking the policies. Usually the absolute minimum for a full-time candidate is 11 or 12 meetings a year i.e. 1 meeting per month. For part-time PhDs, the absolute minimum is half of that. Now this might be sufficient to some topics especially in the social sciences and in the humanities although it is recommended in these disciplines that one meeting every fortnight occurs. Ideally for a full-time PhD candidate in the sciences,  a meeting every week is the most appropriate. The meeting should be at least an hour and should focus on what you have done in the last week or last period, what you will do until the next meeting and covers the supervisor’s feedback on a particular research material or on your progress and any issues you both might raise.

Returning back to the topic of investigating the personality of a prospective supervisor:  you can also use a social media investigation to try to figure out the personality. A lot of academics have social media accounts: Twitter, Facebook, LinkedIn etc… You have no idea how much information you can gather in terms of personality from just looking at social media profiles of people and from posts.  I mean you probably heard that companies, electoral campaigns, governments etc.. are using social media to psychologically profile people on a massive scale via the use of Artificial Intelligence. A lot of people “leak” their personalities, their fears, their insecurities & their psychological complexes in their social media comments, tweets and posts. Per example, if you find the academic in question giving negative comments on Muslims even if they are veiled under an indirect PC language that means a NO NO for you especially if you are a Muslim or Arab or if you find him/her giving bad comments against Jewish people or against any other category or minority that also means a NO NO. Please pay attention to those things! I can study the ego of people from just looking at their social media posts on Facebook. Per example, if a person post on Facebook their acceptance of a certain paper (supposedly in a high-quality conference), this informs me that he/she needs attention (probably some ego issues). S/he cares about appearances so much and seeks grandeur as publishing in conferences is something that any good academic do anyway, in the sense of publishing many papers every year in high quality journals/conferences without the continuous need for the publicity! You got my point!

There are a lot of indicators that could tell you a lot of things, all you need to do is to pay close attention! That’s it! All of this, if taken into consideration, will ease the PhD and make you find with more confidence that jackpot [Smart and Good] adviser that everyone is looking for.

Now one logical question, you might be asking yourself: how the hell can I know all this, if I am on my computer far away from the department or from the academic life? How the hell would I even investigate and ask people? Well!! This is true, you cannot do it easily. This is why I always recommend that if you want to do a PhD in a particular university /school, try to do a 1 year Taught Masters degree first granted that you can afford the time and money (either self funded or via a scholarship) and granted that you have the will to do so. The Masters degree allows you to meet all the staff and to know them well enough. It allows you also to meet PhD candidates, Postdocs and all the lecturers. Furthermore, the Masters is a good opportunity for you to find the right well-balanced and smart supervisor/adviser that you are looking for.

If you are heeding my advice in this regard, when you choose your Masters dissertation topic, make sure that the topic can be continued or expanded to a full PhD with the supervisor/adviser that you have chosen after considerable deliberation or at least make sure that this is the research field you like. There are two main types of Masters in UK: Masters by Research (usually 2 years full-time) and Taught Masters (Usually 1 year full time). Masters by research are like  mini-PhDs, so you need a supervisor from day 1 which means that you will fall with the same dilemma concerning the choice of a supervisor. On the other hand, in a Taught Masters, you have two semesters where you take many advanced taught modules and in the summer, you have to work on research project and you have to write a dissertation with a duration spanning usually 4 months. In this time, you need to choose a supervisor and a topic. Taught Masters are ideal for meeting all your lecturers, and all PhD candidates & postdocs in the department.

People usually encounter a big problem when they finish an undergraduate degree or even a taught Masters degree and then they decide to do a PhD, especially when they did not plan well ahead the PhD itself while they were doing their undergraduate or Masters.  There is a huge gap to jump from a taught degree (such as the undergraduate or Taught Masters) to do a research degree. In addition, when you do a Masters and then go back to industry, you will find it extremely difficult to return back to do a PhD, mainly in terms of finding the right supervisor and in terms of coming out  “miraculously” with many research proposals. This is why I think when you do an undergraduate or taught Master, always think you will do a PhD even when you do not want to. You never know what the future holds! do you? In other words, make sure your undergraduate/ Masters degrees are initiatory for a PhD degree both skills wise and supervisor wise.

If you want to do a PhD and you are outside academia  and you know well & with certitude that a PhD might span 4-5 years. This means you have already accepted the huge penalty of time in your conscious and subconscious mind. In my opinion, adding an extra year of preparatory taught Masters won’t change anything and on the contrary will be very beneficial. The PhD is already a huge commitment in time, health, emotions, sanity and money (whether funded by a scholarship or not) and as I said before, it is not suitable for many people who have different aspirations in life.

Make sure when you do a Masters to aim for a distinction grade (overall average 16.5/20 and above) which will help you secure a scholarship for the PhD.

If you do NOT want to commit to a Masters, it might be up to your luck when it comes to choosing a supervisor. At least find a way to ask about the supervisor a little bit before committing 4 years of your life. You can clearly know the technical abilities of your prospective supervisors/advisers from their academic papers, articles, books, presentations or courses taught etc.. but you will NEVER know their human side other than by the methods mentioned in this section or by similar methods.

How to manage your supervisor/adviser + audio record all the meetings

Once you are in the PhD, you should be the one who manages your supervisor/adviser not the other way around. A supervisor/adviser is a lovely useful human resource and like any resource need management. A supervisor/adviser has no power over you at all, he/she is just a human resource to help you finish your PhD. The adviser main job is to help you gain a lot of research and transferable skills so that you will become an independent academic, a future adviser or a supervisor. Their duty is to give you constructive feedback of substance not just English language feedback or silly feedback. You have of course to have the willingness to accept constructive criticism and to improve yourself. No need for you to be defensive nor offensive when you receive criticism. There are other components in the UK system that judges your progress and competence not the supervisor. They are mainly the yearly progress reviewers and the Viva Voce examiners. In my university, yearly reviewers can not be current supervisors (primary or secondary) and can not be your previous supervisors. This is in order to make the system more transparent and objective. A yearly Ph.D review is usually a small Viva Voce. Please if you are not familiar with concepts and procedures of PhD assessment and Viva Voce examinations have a look at my well recommended article on the topic even at this stage (i.e. I am expecting the readers of this article to be prospective PhDs, or first year PhDs) and don’t forget to consult your prospective university’s policies since they differ between universities.

So it is your job not to allow supervisors to sell you the idea that you work for them for some reason (unless in industrial doctorates) or that they have any sort of power over you, they don’t in reality! Some of them are psychologically sick with extremely bloated egos so pay attention and draw your lines in the sand from the beginning. Of course, as I keep mentioning the whole article talks about normal PhDs not the industrial ones where your supervisor/adviser might be your boss.

In many universities’ policies, a supervisor  has the job of an adviser + minor monitoring tasks. I believe (my opinion) “supervisor” is a semantically wrong term. It is a misnomer. I understand the term used when you work in a factory and say you have a “supervisor”. The folks in the USA call him/her an “adviser” for an intelligent reason. The only thing a “supervisor” needs to “monitor you” is whether you and him/her are having regular formal meetings. This is important to the university and in similar vein it is important for the UK Home Office folks  especially if you are an international student on a student Visa.

In the PhD yearly reviews, reviewers DO NOT have to take into consideration the report of the supervisor if they find that you have achieved a very good progress. It is up to the yearly reviewers to decide what is your actual progress. Usually reviewers take the opinion of the supervisor (s) only on the ground that he or she have worked with you closely and know you better than them but they can ignore the supervisor’s report especially if they find it biased and subjective which is always a sign of an unhealthy & unprofessional relationship between the candidate and the supervisor. You should make sure that you raise any issue  (small or big) with the supervisors first  in order to give them a chance. Afterwards talk to the management of your school such as the Head of School (HoS) or the Director of Postgraduate Studies (DoPG) as soon as possible and even raise these issues in the yearly review session. The opinion of the supervisor is important but do not dictate anything and can be ignored easily in case of foul play or subjectivity.

You need to schedule the formal supervisory meetings yourself and don’t forget to send a short agenda of what will be discussed in the meeting and send a copy of the material that you need feedback on before the meeting. Send these few days in advance so that the supervisor has enough time to read and prepare for the meeting. Never go to a formal supervisory meeting without any preparation. These meetings should be productive and useful. Both your time is previous and that of the supervisor.

Another thing that I STRONGLY advise you to do is to document all your formal supervisory meetings. The best way to document what happens in those meetings is to audio record them.

By doing this: you achieve first transparency, you can use the recordings for archival purposes, for keeping track of what actions should be undertaken by you or your supervisor and sometimes there are too much information and ideas exchanged in a meeting that warrant being recorded. It is good to re-listen to the meetings in case you missed anything. There should be nothing under the table in a formal supervisory meeting and audio recording is the best mechanism to keep things transparent & professional. Many students who I know used the recordings in official university stage 2 complaints and in legal cases to prove wrong doing of supervisors and their complaints were upheld because they have actual concrete submissable evidence.

Always organise your audio recordings into folders with date and time of the meeting. Inside the folder. In my case, I always have a Microsoft Word file containing the minutes of each meeting.

You can of course, send those minutes after the meeting to the supervisor email in this way the supervisor can not tell you that I told you this or that! Actually, I strongly advise you to write the minutes and send them to the supervisor(s).  If you do not want to write the whole minutes, at least send a summary of couple of paragraphs containing what was discussed and any problems raised. A good number of universities even mandate that. These universities require copies of all minutes of meetings especially for the first year PhD review called also the probation PhD review since they can kick PhD candidates out of the program or demote them to Masters. These universities usually have an online system dedicated to record those meetings’ minutes and to upload any meetings’ agendas or audio recordings.

One of my friends in the school had a supervisor who is old. He is really a nice guy but keeps forgetting stuff such as he always claims that he asked my friend to do something in the previous meeting that he did not ask in reality.

Although you need consent for audio recording a conversation when you intend to give it to a third party in the UK and the US (except if you are living in the states of New York, Louisiana & Texas). Any supervisor/adviser who refuses such consent when requested explicitly – especially after you have specified clearly the reasons of recording formal supervisory meetings- is NOT GOOD FOR YOU.  First sign you should change and flee running. Since if the main objective of the audio recordings is transparency: i.e. if you did something wrong or said something wrong, you will be held accountable – there is proof, same applies to her/him. There should be NO reason for not audio recording a formal supervisory meeting. Of course, you must tell them that the recordings are for personal use or for legal purposes in case things went south. This actually forces the insecure supervisor to come clean and keep his/her boundaries. People are afraid to commit mistakes while being recorded which is a good thing. Actually you do not need to say recordings are for legal purposes because the law supersedes anything so it is implied that the university i.e. the employer of the supervisor or governmental bodies could use the recordings as they wish. In case, you were denied consent and you were obliged to keep the status quo, which you should not do by the way – you will regret that! Keep sending the minutes of meetings to the email of the supervisor and in case you had a disagreement or he/she said or did something wrong  such as he/she harassed you, blackmailed you, verbally abused you, asked you to do tangential things not related to your PhD which is very common, did not give you constructive feedback of substance etc…, mention that in the email immediately after the meeting – do not be afraid at all, since this would constitute a strong evidence material for an official complaint and would save you and your PhD if things went south. I come from a law family, so I know very well the motto “the law protects the weak but not the idiot“. In UK, they say: “the law does not protect the dupes“. So please do not be an idiot!

A side note: In the US, federal and many states’ laws differ concerning the issue of recording someone. Some states such as New York, Texas and Louisiana allow you to record as long as only one party in the conversation have consented which would normally be you :-). You can not record a conversation you are not part of: this is literally spying. The rationale behind that is that if your own ears heard something illegal or abusive or racist  or whatever illegal behaviour from a person you are speaking with,  you have the absolute right to submit that to a court of law as an evidence of wrong doing. No need for the consent of the other person even if you give the recordings to a third party. I do not know anything about your specific country, county or state so please please check! You do not want to break the law yourself.

Even if you do not get any consent in the UK, you can still record and monitor the conversation! How and Why? At the time of writing, in UK under “the Regulation of Investigatory Powers Act 2000 (RIPA), it is NOT illegal for individuals to tape conversations provided the recording is for their own use. Recording or monitoring is only prohibited where some of the contents of the communication are made available to a third party” [Taken from BBC article] except when made available for the sake of law enforcement or for legal courts.  If “a person intends to make the conversation available to third parties” except what I stated above in other words to parties that are NOT in law enforcement or legal & constitutional institutions, “they must get the consent of the person being recorded“. Please review the BBC article that I have quoted above or the Act itself. This means that you can record everything either in face to face conversation or in a phone call or skype call as long as you are NOT sharing anything with any third party but only for the sake of evidence of wrong doing in courts. Of course this is very frown upon and out of the academic decorum to record secretly your formal supervisory meetings for personal use only but it not illegal in UK. Nevertheless, I strongly suggest that you ask for that consent explicitly for less headache. There is no point of doing things implicitly. This RIPA Act is put in order for law enforcement officers, the media & journalists and the public (i.e. including you) to help in uncovering corruption in organisations both governmental and private ones and to prevent or detect illegal behaviour and crimes. In the Cambridge Analytica data scandal, the perpetrator was video recorded without his consent. So as a necessary “predicate” in any court you must show that something unethical or illegal was committed or you have a strong legitimate reason beyond any doubt that you thought something was going to be committed and thus you had to audio or video record secretly or explicitly without the consent of the perpetrator. The judge(s) and the jury if unconvinced will grill you about privacy of the individual and about matters concerning data protection act and if you do not have reasonable responses you would have committed a very serious offence.

Policy wise universities should allow explicit audio recording of official supervisory meetings whether consent is asked by the student or not. In other words, the consent should be taken by the universities apriori from all prospective PhD supervisors before the duties of supervision are assigned to them. Audio recording is always better than other types of recording such as written records since an email or a memo could be argued by a clever sneaky lawyer i.e. in other words, things could be argued that did not happen and thus it is more of : “you said this or he said that” – it is your word against his.  Per example, hypothetically speaking if the supervisor said to his PhD candidate, the F* word or the B* and the candidate was not able to audio record that but was able to send an email concerning that. How can we prove that this was actually said by the supervisor? The supervisor can deny that. Audio recording is always the truth – there is no wiggle for interpretations.  If consent of audio recording is refused the academic should NOT be a supervisor in the first place. In other words, I mean  as an academic if you want to supervise postgraduate/undergraduate students, you have to accept being recorded in official supervisory meetings for transparency purposes under defined conditions such as not sharing them on social media or other platforms or to third parties and to only sharing them with the supervisor, the university, and legal institutions in case of wrong doing.

I have an article which goes into much detail of the forms of abuse, harassment, blackmail, exploitation, neglect, favouritism, quid pro quo among other forms of academic misconduct committed by UK supervisors and academic staff which you can read here. The article provides techniques to protect yourself legally. It is still a protected article with a password since many people are adding material form their experiences. Also there are many permissions that I am waiting for in order to publish the article.  Also few of the folks  who legally deal with  investigating cases submitted by students are still adding valuable advice to the article. Students do not know that they are in power and that they are protected and I blame UK universities for not giving them mandatory seminars explaining both their rights & their duties (not only their duties!).

I can say a common form of blackmail is when a supervisor wants to end supervision because you disagreed academically with them or refused to do a tangential or irrelevant work or refused to put a name of an academic who did not contribute to a paper of yours. The latter is known as abuse of power.  Students do not know that this is a form of blackmail that can put the supervisor in very serious trouble from a UK law perspective and is unethical academically speaking. If you did not do anything wrong yourself, he/she can not end the supervision without serious repercussions on his career. Blackmail is a very serious offence in UK law which could  lead when indicted to a maximum of 14 years of imprisonment.  The most important  matter for any case is that you should  ALWAYS ALWAYS have strong concrete submissible evidence. So make sure that you do that from the beginning. You never know! Emails and audio recordings are the best evidence.

Continuously ask the supervisor for feedback on your progress: The majority of sane people think that the supervisor is the one who should tell you how they are doing! or should warn you if their progress is not good and that is true and logical. If he/she does not say anything, for the majority of people with a high IQ, this means things are good. Since the system calls them “supervisors” after all. You see the wrong semantics used in UK. In addition, the majority of people around the world consider silence to mean acceptance or agreement (everything is OK). A supervisor has to tell you and warn you when things are not working or when he/she thinks are not working (his/her judgement might not necessarily be true) so that you have the chance to rectify in due time. It is unethical to do otherwise. Guess what, some UK supervisors don’t do that and then they pour it in the PhD review report at the end of the year. This is literally gutless, cheap and unethical behaviour. We come from societies where we like people to look us in the eyes and tell us their opinions.  I strongly advise you, at least every month or two, to ask the supervisor clearly about your progress and have it recorded or mention it in the minutes that you send to the supervisor’s email. I used to ask my supervisor to give me a progress colour (Green, Yellow, Amber, Red) because our yearly review system gives PhD candidates “colours” as a way to grade the progress in their PhD. This makes you aware of your progress in order to rectify in a timely manner and then it makes it clear for the yearly reviewers that because you have disagreed with supervisor about X issue on date Y, the supervisor is saying your progress is not good while it was good before this “disagreement”. Trust me! it happened to a guy in our school.

How to choose a good PhD topic that does not cause you serious mental health problems?

Justin Zobel lovely book titled “Writing for Computer Science” has a big part of a chapter that talks about how to choose a topic for research for a Master or PhD. You can find a lot of great ideas there. The ideas are NOT just for computer science and I believe they are applicable to any discipline. I am avoiding repeating them here but I will mention few notes. You usually approach an academic that is doing research in what you are interested in and then you would work with the academic in question (i.e. the prospective supervisor) on a research proposal.

My advise to you is to always always always choose a topic  that:

  • [First]  is in tune with your own capacities & skills so far. Don’t shoot your leg! Don’t be rambo!
  • [Second] you like very much or that you are passionate about. I know this is sometimes not easy to have since you might be tied to a topic for funding purposes. Bare in mind, you will be stuck with this topic for many years. You will not succeed if you don’t like what you are researching!
  • [Third], it should be substantial and challenging enough to be considered at a PhD level.
  • [Fourth], there should be at least one person who can and who is willing to advise you on it in the school or department. Ideally it is better to choose a topic where there are more than one person with the expertise in the department or the school.
  • [Fifth] ideally, it is better to choose a topic that allows you to join a complete research group where you can have a lot of feedback and help from postdocs, PhD candidates and other researchers and where you can collaborate with others.
  • [Sixth] Choose a topic that can help you get an academic or industrial job after you finish. You need to plan for the period after the PhD! Ask yourself: can this topic facilitate getting me postdoctoral positions, or can this topic facilitate landing on lectureships in prestigious universities or even on jobs in the industry? Another question people ask especially for scientific or technological research: can I transform the research done on this topic into a commercial product? PS: When you ask yourself these questions,  search around online or ask conoscenti people to find out the answers!

Substantial contribution to knowledge is a major requirement for a PhD but “substantial” does not mean in any way splitting the atom. 🙂 The topic has to be “doable” and in the time frame of a PhD. Do not choose very ambitious or very exaggerated topics! These topics are a recipe for failure since they will usually take a lot more time than a PhD.

There are two essential requirements for a PhD here: the scope requirement  and the original contribution to knowledge to the field requirement.  Please read my article on PhD requirements, assessment and Viva Voce preparation to have an idea of what I am talking about here.

Pay attention: A topic that is “revolutionary” in a PhD is very very very rare and in the absolute majority of cases where students choose such a thing end up literally failing. You can not split the atom in a PhD 🙂 , please don’t do that for a PhD. Do it afterwards if you can 🙂 Well, you can of course ignore this advice. I can not prohibit you but in my experience this does not work and has never worked with a lot of students. After the PhD, it would be OK to choose a topic that leads to a Turing Award or Nobel Prize. This is especially true for the average disadvantaged Joe coming from a poor country in Asia or Africa who don’t even have till today a full 24 hours electricity coverage.

Ways in which ideas worthy of a PhD investigation are born?

  • Conversations with a lot of researchers in the field about an observations you have or a problem that you heard about or that you read in a survey or about research gaps that are identified etc… This kind of brainstorming and conversations lead in many cases to interesting & worthy PhD topics. Usually either you have an idea or your prospective supervisor has one and you both end up discussing it and expanding it. You will normally develop a first draft of a research proposal which the supervisor can give you feedback on it and help you to hone it further and further before submitting it with an application.  You need in advance to send your transcript grades, your CV, your GitHub URL, your academic papers among other indicators to a prospective PhD supervisor so that he/she know that you are a worthy candidate to spend time with and to help with a research proposal. Contact some references from previous degrees and let them send reference letters to the prospective PhD supervisor. This helps a lot. The academic world is very small so your prospective supervisor will probably know your previous lecturers and professors.
  • Survey papers are amazing in summarising, contrasting and presenting research studies done in a certain topic. All descent surveys expose the challenges and problems that remain to be solved. These are great ideas to choose from. I mentioned different places where you can find published surveys in my article covering writing literature reviews.
  • Previous PhD theses of the school of your target are usually shared online and all PhD theses have further research sections with tons of ideas. It is good to contact the person who wrote the thesis in order to discuss those ideas. A lot of people are happy to engage in such conversations even if such person is not your prospective adviser. There are also many ideas worthy of investigation in future work sections of academic papers.
  • … there are many other ways but for brevity 😉 Yeh right! I will move on …

Whatever topic you end up choosing, always start in a simple way i.e. in simple measurable steps and begin to expand incrementally on that. Adopt SMART goals along the way!

Now moving on with the skills…

B – Master Statistics

I know! That one course or 2 courses, you have taken in the undergraduate degree. Some students even do not take any statistics courses in their undergraduate especially educational systems that follow the American credits system where you can opt out easily of many courses and still graduate. Statistics will surely hunt you in the PhD! Well, it will hunt you when you do any type of research!  My advice to you is: learn statistics well and I mean very very well!!! PS: There are very few research fields and topics (especially in the humanities) where you might not necessarily need statistics.

In Computer Science per example, the usage of statistical tools such as correlation, regression and hypothesis testing is very common in research[5]. So you need to master statistics because there is a big probability that you will end up using many statistical functions and tests. In similar vein, you need to able to understand statistical calculations and tests that are used in tons of academic papers and theses that will definitely read during your PhD. No sane and intelligent person will argue against what I recommending here.

I strongly advice you to do a very good refresher on Statistics (First and Second Level – 2 semesters courses as an absolute minimum).  MASTER hypothesis testing and statistical tests at least the essential and the most commonly used. You should understand the theoretical background of each test, the assumptions necessary for the test, what research scenarios (type of variables…) are ideal for the test, what research scenario in which you should avoid using the test:

  1. Tests of normality of a distribution: Kolmogorov-Smirnov test, Shapiro-Wilk test
  2. Correlation Tests: Pearson Test,  Pearson Partial Correlation, Point-Biserial Correlation, Spearman Correlation, Kendal’s tau-b, Goodman & Krouskall’s gamma, Somers d, Mantel Haenszel test of trend, Cochran Armitage trend test, Fisher Exact Test, Chi Square Test  and many others…
  3. Prediction Tests –  All Regression tests: Linear Regression, Standards and Hierarchical Multiple Regression, Binomial Logistic Regression, Ordinal Logistic Regression…
  4. One Sample Tests: One-sample t-test, Chi Square goodness of fit test…
  5. All types of ANOVA (Analysis of Variance) tests (One-Way, Two-Way, Three-Way… ), ANCOVA (Analysis of Covariance), MANOVA (Multivariate Analysis of Variance)….
  6.  Distribution-free tests such as Wilcoxon-Mann-Whitney test
  7. Useful non-parametric tests such as Mann-whitney U Test, Wald-Wolfowitz runs test, Wilcoxon signed ranks test, Kruskal-Wallis one-way analysis of variance by ranks test, Friedman’s two-way anaylysis of variance test …
  8. …. many many more…

Learn and understand how to draw at least the following plots, when to use them and what they actually show you:

  1. Plots that shows distribution:  histograms,  box plots a.k.a. (Box and whisker plots), Mosaic Plots, Violin Plots, Density plots, Ridgeline plot or Joyplot, Coplot (famous in Psychology)…Cumulative Distribution Function (CDF),  Probability Distribution Function (PDF)…
  2. Plots that shows correlations: scatter plots & scatter-plot matrix, line plots, heat maps (used so much), Correlograms, Bubbles plots, density plots …
  3. Plots that shows Ranking: Bar Plots, Spider Plots, histograms, Word Clouds, Stem and leaf plots…

Box plots are one of my favourite charts. They show you a lot of information about your data. They are very good at showing distribution but in a similar vein, they also show you outliers (meaning extreme values), the upper adjacent value or in case of no outlier the maximum value, the upper quartile, the median, the lower quartile and the minimum in case no outliers or what is called as lower adjacent value.

One of the great things that I have learned from my supervisor is to never use a Pie Chart. If you want to know why please see this article and this article. It is frown upon in academia especially in natural sciences and in Computer Science even in social sciences. Pie charts are for managers on the slides of a presentation of a Monday morning. Pie charts are used for public consumption like for a TV show since they are very simple. Pie charts should never be used in any descent scientific research paper.

For all the statistics that you learn on a theoretical level or that you revise again for the sake of PhD preparation, I strongly advice you to make sure that you know also how to do all these statistics on a practical level. I mean learn how to do hypothesis testing and how to draw different types of plots in a specific statistical package or using a specific statistical programming language.  Many PhD candidates use either the R statistical language, SPSS,  Matlab, or Python data science modules. Kindly refer to Skill C.

Recommended books to read on statistics:

  1. Discovering Statistics Using IBM SPSS Statistics Fifth Edition by Andy Field. This book teaches you theoretical statistical concepts and SPSS in the same time and it does that in a funny way. The author has a sense of humour. Another version of the same book that covers R: Discovering Statistics Using R. Don’t worry about R and SPSS, I will explain them in the next skill.
  2. Statistics in Plain English by Timothy Urdan
  3. Statistics by Hays.
  4. Head First Statistics by Dawn Griffiths – actually I am in love with the Head First series and the statistics book did not disappoint me.
  5. 100 Statistical Tests by Gopal Kanji – details of a lot of common statistical tests. Very easy to read and ideal to consult when you need to learn about a particular statistical test.
  6. Oxford Handbook of Medical Statistics by Janet and Philip Peacock – this book is a very valuable guide covering a range of statistical techniques and definitions.
  7. Problem Solving: A Statistician’s Guide by Chris Chatfield.
  8. Applied Multi-variate data analysis by Everit and Dunn
  9. Multivariate Data Analysis by Hair and Anderson.

Recommended Books to read on plotting:

  1. The Visual Display of Quantitative Information by Edward R. Tufte – a classic that MUST be read by any PhD candidate if she/he is planning on drawing even a single plot. This book is recommended by many PhD supervisors.  It is not important to visualise your data for the sake of visualising. You plots needs to be needed first of all. They need to be to the point and with minimum ink on the page 🙂

Statistics and Research Methods books

This category of books is very informative since you learn essential  descriptive and inferential statistics and in the same time you learn most important research methods in a certain field. There is another skill explained later in this article which emphasises on the need & pertinence of having a strong knowledge in the research process and in the research methods of your field/topic before embarking on a PhD and I strongly advise you to read dedicated books on research methods in your particular field of research before you start your PhD. For brevity and just to give you only one example discipline: the following are few examples of books that contain both research methods & statistics needed in Psychology. You can find many similar books for many other fields.

  1. Research Methods and Statistics in Psychology by Hugh Coolican. This book covers both quantitative, qualitative and mixed methods research. In addition, it covers a lot of statistics needed for Psychology research.
  2. Introduction to research methods and statistics in Psychology by Ronald McQueen and Christina Knussen – although this book is more targeted to undergraduate research than to postgraduate but nevertheless it is amazing, rich and easy to read (i.e. many research concepts are explained very well). It is also very useful  for an introductory level for research at a Master or PhD level. It is even very useful for research outside Psychology.

Web resources

  1. The Statstutor website is a very good web resource for learning statistics. To get you started, I advice you to read the Statistics Tutor’s Quick Guide to Commonly Used Statistical Tests .
  2. The lovely University of Glasgow STEPS Statistics Glossary which contains a wealth of resources.
  3. If you are into free e-books and do not want to spend money on hardcover/kindle books, I got you covered – Check out the CAST collection.
  4. WhatTest website is aimed to help out researchers and students to decide on what to choose as appropriate statistical analysis – Marvelous!!! Amazing guides and to the point!

Online Video Courses

Udemy platform

The following are some of the recommended online video courses to just get you started!

  1. Statistics for Business Analytics A-Z
  2. Become a Probability & Statistics Master
  3. Workshop in Probability and Statistics
  4. Statistics for Data Science and Business Analysis
  5. The Simplest & Easiest Course on Hypothesis Testing
  6. Math for Machine Learning (you know machine learning is nothing but statistical learning)- this course is more for Computer Science students interested into the Math of Machine learning

Interactive Multimedia Online Course

  1. Online Statistics Education – a very good multimedia course that teaches introductory statistics.

Suggested YouTube Channels and Courses:

  1. Crash Course Statistics: This course is amazing. It is a good crash course refresher on basic statistics. Actually all the CrashCourse YouTube channel is amazing. You should subscribe to it.
  2. MarinStatsLectures YouTube Channel: all sort of good statistics lectures. Just amazing!
  3. statisticsfun YouTube Channel: all sort of good statistics lectures. Simply amazing!
  4. BrunelASK Videos on YouTube – a collection of short videos covering statistics (including tests) in SPSS and in Excel.

Side Section – Online Video Courses Platforms

Before I proceed with the skills, I wanted to create a side section about online video courses’ platforms or MOOC platforms and talk about their importance these days since I know a lot of people are not yet aware of them or of their pedagogical benefits. Not to mention, some of them are very democratic, meaning if you know a skill in demand you can create a course and begin to earn a lot of money. 2 instructors who I know personally made from several video courses over £500000 in revenue. Imagine! It is simple math –  some courses have hundreds of thousands of students enrolled in them from all around the world. It is not a bad idea to consider to teach a topic related to your research on such platforms after you finish your PhD or even during your PhD.

We are in the age of streaming massive online courses from platforms such as Udemy, Udacity, CourseraedX, Pimsleur, Khan Academy , , Codeacademy, Alison, Tuts+, Open Culture, OpenLearn, FutureLearn, Skillshare and to your computer or to the mobile device in your pocket!! A lot of these platforms offer numerous free online courses and even free nano-degrees/micro-degrees, specialisations in many technologies, skills and disciplines. Many courses are given by professors from very reputable universities. In addition, you have also your good old YouTube and its numerous free educational channels!  iTunes U used to be something big but its light faded for some time now.

I lived all my life in a situation where  the only option for me is to read a complete book in order for me to acquire a particular skill or I had to read snippets of articles online that gave me incomplete and sometimes wrong way of learning something. That time fortunately is gone! Who has the time anyway to do that? I used also to be obsessed by pseudo sciences like Speed Reading and PhotoReading out of necessity of course. Both do not work at all by the way. You would think that they do work for a brief time but then you return back to reading habits. I wanted a way to read faster so I can reduce the amount of time needed to acquire the particular skill needed for a project and then finally to be able to begin to work on the actual project!

The way that I  approach the matter is as follows. Suppose I want to learn a skill X, I would browse for a complete course or 2 courses  that are either free or even paid- depending on my needs. Then I take that course spending 2 to 3 days watching the course. Such courses might be introductory.  I like to watch the course even the ones with 40 hours or 50 hours of video material like I watch a Netflix or Amazon Prime series meaning not spending time doing stuff but getting over with the material and understanding it very well. This is because I am experienced now. I already know many programming languages so if it happens that I want to learn a new programming language, I can use an online course for gaining quick familiarity. This is my learning style. It does not suit all people. After taking the video course (spending 2 to 3 days max). I then begin straight away with the project at hand and if I need any further pointers or any further learning I can go online or even buy advanced books if needed. If I got stuck on a technical/programming issue with no material online to cover it, I would ask questions on platforms such as stackoverflow  or similar places. There is the Stack Exchange Network which is set of Q&A online platforms where you can ask questions about different topics. Stack Exchange Network covers many topics pertaining to many fields such as technology, programming, statistics, life, politics, law, religion, history, art, finances among many other disciplines.

It is pertinent to mention another good reason to enroll in platforms such as Coursera, Udacity and Udemy which is the fact that all of these online platforms have mobile applications on Android, Microsoft and iOS operating systems, which allow you to view their courses material while you are on the go (on the bus, on the train etc…). Many of these platforms also allow you to download the material to be viewed completely offline which is quite handy in the absence of an Internet connection.

Udemy Video Courses

Udemy is cool! Why? because with just 9.99$ or 9.99£ (it depends on the currency of your country), you can buy a 40 hours video course covering SPSS or R or on whatever topic you can imagine. You can buy a 70 hours course covering the Unity game engine!!! How cool is that? Do not buy any course when you see a price higher than 10$. Udemy is a bazaar market for courses. There is always a reason for Udemy to lower the prices to 10$ (St Patrick day, Summer sale, Black Friday, Christmas, Fathers Day, Mothers day, New Students Sale, Beginning of the Year Sale… I mean it is quite funny… not to mention that a lot of the instructors on the platform can give you discount coupons that would reduce the price of a particular online course to just 10$. I bought a very good and informative 20 hours video course for only 2$ by using a coupon provided by the instructor. There are many social media groups on Facebook, telegram among others and there are also many websites that offer free coupons for very good Udemy courses.

Udemy has mobile apps for Android and iOS, where you can download the video lectures and watch them offline. The Udemy platform provides instructors with the ability to write quizzes and allow instructors to interact with students through Q&A forums, give them quizzes and programming exercises among many other features.

Udemy is one of the most democratic platforms that I have seen so far. I touched on this fact before. You can become an instructor and teach a topic that you are proficient in and enjoy teaching it while earning good money due to the sheer size of students from around the world! You only need a very good microphone (to avoid noise), a laptop or PC and a screen recorder software like Camtasia (cost money) or OBS Studio (free & open source) which are very famous  and voila! Some famous instructors even create complete studios with acoustic panels and lights setting for online teaching.

The disadvantage of Udemy is that many times you might encounter very terrible courses with very terrible instructors. That’s what a bazaar or souk of courses is about! It is like those souks that you go to in the middle and far east,  and where many times you actually buy so cheap a lot of good and valuable  stuff such as buying a 500$ antique armoire with only 40$!   or you can be screwed by buying terrible things and wasting money. Normally, online courses that are best sellers & highly rated are very good in general. Pay attention also to the number of students enrolled. Always keep an eye on the reviews of the courses, especially the negative ones. Generally, I always read negative reviews before buying anything online.


Pay attention to the fact that a good number of employers don’t respect certifications of completion obtained from many MOOC platforms such as Udemy simply because anyone can watch the videos or simply mark them as complete and anyone can do the quizzes or mark them as done and by doing that will get the certification of completion. This is why employers might prefer to see certifications from online MOOC platforms such as Coursera since first, you are taking a course given by a reputable university on a specific academic topic and second, the quizzes and the assignments are changed periodically and are usually audited for quality. So please bare this in mind.

Monthly fee based video online platforms such as Skillshare and Pluralsight

Skillshare is like the Netflix for online video courses. You pay only a small monthly fee instead of buying each course alone. Skillshare provides, at the time of writing, access to tens of thousands of courses. Other similar learning services are, and  but they are more oriented to Computer programming. You can find a lot of good material on these platforms. What differentiates Skillshare from other similar platforms is that it is democratic like Udemy meaning that you can find both bad courses and good courses. In addition, you can teach a course on Skillshare.

The beauty is in the fact that you pay only one fee which could be very low. It depends on what offer you get. You can pay as low as 7 sterling pounds a month with access to tens of thousands of courses. For readers from the future – this will probably change.

There are also many free courses if you do not want to pay a premium monthly fee but these are usually very short or very terrible. Skillshare has iOS and Android mobile apps so that you can learn on the go.

Bare in mind, there are a lot of topics not covered in Skillshare at the moment which you might find equivalent courses covered in other online video platforms such as Udemy, Coursera or Udacity but this is bound to change in the future.

PS: some academic institutions have subscriptions to such online services especially to Linda so please check before you subscribe and waste money.

C – Master 1 or more Statistical Packages or Statistical Languages

This is a very important skill needed for a PhD. Even if you are old school and you know all the math behind all the statistical tests and you like to do them manually. You need to master a statistical software or package or a statistical and plotting language since we do not live in the 60s.

You strongly adviced to learn at least ONE of the following statistical tools/ packages or languages stated in this article. Now you ask which one? I can NOT give you a preference since this depends on the field. Although I tried to specify the pros and cons of each software package, statistical language or plotting language.

As a good advice here, ask PhD candidates in your particular field about what statistical packages or languages that they have used heavily in their research or ask academic staff that you know or ask your prospective adviser. That has been said, I asked many  PhD students from different fields (both in social sciences and in natural sciences) and it seems many of them either use R (top position) or SPSS (second position), or Python or MATLAB. These 4 seems to be the most famous kids in the neighbourhood 🙂

Knowing MATLAB and knowing how to write scripts in it seems to be crucial with many folks doing research in many fields in Biology, Biotechnology, Neuroscience, Medicine,  Mathematics and Psychology. don’t ask why, I do not know but these results comes from my friends sample. MATLAB has a lot of tools out of the box for many fields such as Mathematics and Statistics, Control Systems Design and Analysis, Signal processing and Communications, Image processing and Computer Vision, Computational finance, Computational biology, application development, database connectivity among many others.

Realise Microsoft Excel is not on the list that will follow. Microsoft Excel is a powerful spreadsheet software but not quite suitable for a lot of cases where you need certain niche academic visualisations or certain “statistical tests”.  You can use of course Excel for that but you have to do them manually if you know the Math. Learning  R or SPSS or Python  is a very good investment.

Microsoft Excel plots are easy to do and I should say Microsoft Excel evolved to a stage where you can create very beautiful plots like heat maps and box plots but they do not give frankly a professional look especially in the context of sciences papers and theses. I can spot an Excel or similar software chart from 50 km away. Nevertheless, Microsoft Excel with Pivot Tables and Visualisations might be an amazing skill to add to your skills set but not so much for a PhD sake, maybe if you are in some fields such as Finance or Management. But again PhD candidates in these fields, tend to prefer SPSS and R over using Microsoft Excel or in the best cases, they usually demote Excel to only very specific tasks like data cleaning (a.k.a data scrubbing) to be used with SPSS or R or SAS.

Don’t get me wrong, I love Excel! I taught courses on Excel. I know every function and formula under the sun. I gave full day workshops on Excel. I taught writing Powerful Macros and VBA in Microsoft Excel and Microsoft Word. It is amazing what you can achieve with Excel! Actually, Excel has a funny dichotomy: It is in the same time, very simple and extremely powerful.

PS: Microsoft Excel can do few of the common statistical tests  and hey! maybe this is all what you might need for your research.

For a good Udemy course, I suggest the following:

  1. Statistics for Data Analysis Using Excel 2016 – It is a very good Udemy course. You would be amazed of what you can do in Excel (visualisations, descriptive statistics, t-test, chi square, ANOVA etc…)

Few online resources and published books:

  1. Saunders et al. Book on research methods in Business has a very good but very old Excel guides that can be found here. The statistics you learn still apply to newer versions of Excel.
  2. Excel Statistics: A Quick Guide by Neil Salkind.

If I am you, I would not bother reading books on Excel Statistics before a PhD, I would invest in learning doing statistics & plotting in a language i.e a scripting statistical language such as R or using Python Data Science libraries or SPSS or SAS or similar tools/languages that I will explain later. MS Excel is good only if you know for sure that you will be using very little statistics in your PhD.

The following section is an exposition on why it is important to choose to learn a Statistical & Plotting Language  more than just learning to use a Statistical Software with a graphical user interface.

Important Notes on Scripting plots and Statistical Tests

Before we proceed with the list that shows statistics software or statistical languages or plotting languages that a PhD candidate should master,  I want to state the benefits of learning and mastering writing statistical and plotting scripts that you usually can do with statistical and plotting languages OVER using statistical software. When you write the statistical procedures and plots in a script (i.e. statistical code), you are able to “run” or “execute” the script whenever you want, wherever you want, and to produce with the same or with other data-set, the plots or statistical tests that you need. This guarantees reproducibility which is one of the major tenets of doing science and research. In addition, scripting allows you to re-utilise your code for other scenarios that you might encounter later in your research.

Benefits of scripting which are also benefits of statistical/plotting languages over graphical software packages:

(1) Scripting creates a unified tidy appearance for all your plots (labels, axes..) in your papers and in your PhD thesis and this is a plus.

(2) The most important aspect of sciences is:  reproducibility. Scripting helps you tremendously in achieving reproducibility. Actually as a PhD candidate, you should keep all your underlying data with the statistical scripts  that you have created in a folder or some kind of organisation and set them aside to be re-used by yourself in another project, or by others in your research group and even to be submitted with your thesis or put online so that they can be available to others to replicate (on GitHub or Bitbucket or sites that share data-sets & statistical code). This is a big plus! Have a look at the Zenodo platform for archiving datasets, statistical scripting and programming code.  Zenodo is a research archive for datasets and for programming code that underpins digital outputs of research. It is purposed to make these digital outputs “citable” academically speaking. It gives you a DOI for your code repositories.

(3) Scripting is the recipe that tells you the story of how a plot came to be and how a statistical test was done. If you are reading this article before you start your PhD, I advice when you to learn  a scripting language and to create templates for all common plots and tests. Per example, an R template or a Python Template or a Gnuplot template for creating a box plot, or a matrix of box plots, or a heat plot, or a mosaic plot or an R script for doing an ANOVA test vel cetera.

(4) Usually plots produced by non specialized statistical tools look pixelated or just weird. There is a simple explanation: Common Computer screens operate normally at 72 dpi (dot per inch) resolution, printed material can demand at a minimum 300 dpi and even more reaching 600 dpi or even higher to look good. Consider the example of a big poster with a graph on it. The only way you can achieve such very good resolutions of images worthy of printing, is either through few specialized graphical statistics software packages which can have such feature or more commonly using statistical and plotting languages which all have such capability. R or Gnuplot per example, allow you to spit out (export) that kind of very high resolution PNG, JPEG or PDF files or whatever extension you need for your academic papers or for your thesis. You can of course use scalar vector graphics plots or coded plots like TiKz and PGF for LaTeX. In either way learn how to produce beautiful high resolution plots.

(5) The benefit of separation of data from appearance: Plotting languages such as Gnuplot (will be explained in detail later), and statistical and plotting languages like R and specialised Python data science libraries have the advantage of separating the appearance i.e. how the plot looks like from the actual raw data and the cleaning process that was done on it.

(6) Statistical and Plotting Languages (ex: R, Gnuplot…) are by nature more powerful then Statistical Software (ex: SPSS) especially if they have a big community of users and developers. Of course these languages are not easier to learn. This is for a very clear reason that if there is a need for a plot  or statistical test  in statistics, there are definitely people who wrote a module or package to do that which you can import easily in your scripts. This might not be available in graphical user interfaced Statistical Software especially the commercial ones such as SPSS where you need to wait for them to include a feature or a plot or statistical test that you might need right now. Something that might never happen even!

Important Notes on Plots in general

(1) Plots are essential instruments that show us what the data is actually saying from different perspectives and angles.  It is very hard for you as a beginner to know what type of plots suits what type of scenario, what type of data and what type of relation you are trying to show between variables. I touched on that before, per example, use only line charts when showing some sort of relationship between variables.

(2) Graphs will be always printed in white and black: So try to use a legend and a plot that is understandable to the reader when printed in white and black. I love using colours in graphs. If you are like me, make sure you to differentiate lines in charts by dots, dashes, crosses etc…. I don’t always follow this advice myself.

(3) Consistency in labelling, axes, legends: Axes should start generally from 0. There are some other cases where this is not needed. Labels and axes font and size should be legible and tidy.

SPSS (Statistical Package for the Social Sciences)

SPSS is extremely powerful software package and I strongly advice you to learn it and to master it. Not only you should learn the tool but learn how to do every possible statistical test with this tool. Learn also how to draw all the plots mentioned in the previous section. Please do not be fooled by the word “Social” in it, it can be used by the natural sciences folks. A lot of PhD Candidates in Computer Science use it.

SPSS can do amazing and powerful tasks with couple of clicks. When you are doing a PhD, you need every minute and hour so no need to spend writing complicated R  or Python scripts, if you can do what you want to do with three or four clicks in SPSS.


(1) Easy to use once you learn how to use the software, how to plot common charts and how to do statistical tests. The learning curve is easier than learning R language  or Python statistical modules. This is why it is preferred more by social sciences folks.

(2) Powerful to a point: But hey! maybe that is exactly the power that you need for your research. Actually a good amount of research done in scientific, social and humanities fields need just the power of what is available in SPSS. So you are safe! You can of course generate scripts in the language of SPSS so you can run your scripts again when you need to do the same statistics.


(1) Cost a lot of money (not free). You need to have a university or educational institution license to work with the software. Well!! this is contradictory to your current situation now since you did not start your research degree yet.

(2) Not powerful enough. A lot of types of plots or advanced statistical tests are NOT available in SPSS but are available in languages such as R  or Python statistical libraries. In R, you can do machine learning, deep learning, advanced data analysis. It is a complete programming language which is extremely powerful.

A cool Web Site Resource for SPSS folks Called Laerd Statistics

Laerd Statistics: is an amazing website which contains literally everything you need to learn about statistical tests in SPSS. They have also material for STATA. The site contains the necessary assumptions for every test you need to be aware of.  Even the web site contains how to write or how to report your results following different styles such as APA among others.  In other words, the web site teaches you the right academic way to write your SPSS statistical results. There is a lot of free material but the premium material is more complete. The 6 month subscription is only few pounds ONLY (the price of a cup of coffee).

Furthermore, Laerd Statistics web site contains a Statistical Test Selector Tool which is a kind of an expert system that tells you what statistical test you should use for a particular case. You give it the type of study and variables measurement types and it tells you what are the most suitable statistical tests to use.

Recommended SPSS Books

When I recommend Books I don’t just  throw at you some fluff books that lead you nowhere The ones that require a century for you to begin to grasp what you need. I hate the books that I need to read a 100 pages before I begin to read the meaty part. A lot of authors just like padding. I recommend normally books that are enjoyable to read and lead you somewhere after reading them. Books that have tangible effect. I read many books that wasted my time before.

The easiest best book for Learning Statistics and then applying it using SPSS (Top ONE) is:

  1. Discovering Statistics Using IBM SPSS Statistics Fifth Edition by Andy Field.

Pay attention please, the above book is not an SPSS cookbook. It teaches statistics as a main focus with applications in SPSS. You have many books that teaches only SPSS to an advanced level.

Actually, you have two types of books. First Types: Actual Statistics books + maybe a limited explanation of some software or language to apply what you have learned. The second type involves Statistical Software/languages Cookbooks. You normally buy books from the second category only when you already mastered Statistics concepts in general not the other way around like a lot of people do and who end up nowhere.

Other good SPSS books:

  1. SPSS Survival Manual: A Step by Step Guide to Data Analysis using IBM SPSS by Pallant.
  2. Quantitative Analysis and IBM® SPSS® Statistics: A Guide for Business and Finance by Abdulkader Aljandali

Some books are for certain fields such as Psychology. I recommend the following list of books based on recommendations of a friend of mine who is doing a PhD in Psychology :

  1. Statistics Without Maths for Psychology by Prof Christine Dancey and John Reidy
  2. Making Sense of Statistics in Psychology – A second level course by Brian S. Everitt. This book although old but amazing. I even read couples of chapters from it and learned a lot.
  3. An Introduction to Statistics in Psychology  by Dennis Howitt and Duncan Cramer. If you are doing Psychology, I envy you for the existence of this book!!! Everything you need to know from descriptive statistics, significance testing, t-tests, chi square, analysis of variance and many many more. I read the first 2 parts of it and I will tell you this: the book is great even for us in Computer Science. Don’t forget that computer scientists do a lot of experiments that involve human subjects.
  4. Statistical Methods for Psychology by David Howell.

SPSS Cookbooks Just for Psychology:

  1. SPSS for Psychologists by Nicolas Brace et al.

Suggestion of Good Udemy Video Courses for SPSS  – some of them are recipe courses not purely statistics courses:

  1. Statistics/Data Analysis with SPSS: Descriptive Statistics
  2. Statistics / Data Analysis in SPSS: Inferential Statistics
  3. SPSS For Research
  4. SPSS Masterclass: Learn SPSS From Scratch to Advanced
  5. Statistics Made Easy by Example for Analytics/ data science

Stata Software Package

Stata is a statistical software package used for applied research. It provides different capabilities such as data manipulation, visualisation, statistics and reporting. It has also a very good integration with the most common database management systems. Stata is not really commonly used at least from my own experience. I have met during my PhD a very large number of PhD candidates from many disciplines and I usually like to ask all candidates that I meet about what tools they normally use. I never came across someone who has told me that he/she is using Stata. Now of course I might be wrong but anyway I included it here for completion sake. I advise you to ask previous or current PhD candidates/postdocs in your prospective research group or your prospective supervisor about what statistical software package or statistical language you might need to use in your research.

SAS (Statistical Analysis System)

SAS is both a statistical software and a statistical & plotting programming language that is extremely powerful. The software is NOT free but there is an educational version suitable for students. You can with SAS (as software and as a language) handle very large datasets, scrub your data, plot it with different visualisations, do statistical tests and techniques and much much more. SAS is almost equivalent to R in terms of being a good choice for you in your PhD. I said ‘almost‘ since R is a lot more powerful especially when it comes to machine/deep learning and advanced ‘edge statistics’.

R is the open source equivalent if you want of SAS. Another analogy if I am allowed to make, R is like Linux OS and SAS is like MS Windows. SAS is easier to learn and use compared to R.

I wrote few scripts in SAS and compared them with same procedures that I wrote in the beginning of my PhD in R. From my experience, R scripts can become very messy, very fast. On the other hand, R can do stuff that SAS can never do especially in aspects of predictive models, machine learning and deep learning. R also due to its free and open source nature has always packages with the latest statistical techniques, something you might not find in SAS. That has been said, you do not have to worry about that ‘edge statistics’  that you find in R if you do not need it in your research.  But to be fair, R is more valuable to you in my opinion, if you want to learn one or the other.

Tableau Software

Tableau allows you to create very beautiful interactive and non-interactive visualisations of your data. It is used mainly in the domains of Business Intelligence and Business Analytics but recently expanded to more research oriented domains. Bare in mind there are many differences that governs the semantics of both the term ‘Analysis‘ (dealing with past data, decomposing it and examining it and extracting meaning out of of data) and ‘Analytics  (dealing with the future: with predictions based on Analysis of past data).

Minitab Statistics Package

Minitab is a commercial statistics package. This is loved by the Psychology folks in particular and by social sciences in general.

——–For More Power———

Go and learn a statistical programming language. All the software packages mentioned above can NOT DO many things that the statistical Languages and general purpose languages can do.

R Statistical Language

R is the king of all statistical automation. It is statistical & plotting programming language. Actually, it is a fully fledged programming language. People use it for all sort of things. It is free and open source. R can do stuff that SPSS/SAS can never do. But that comes with some expense: a learning curve.

If you decide to go with learning R. You should not only learn the language but you should also learn how to do every possible statistical test in it. In addition, learn also how to draw all the plots mentioned in the first section. One of the most famous software/IDE used in R is RStudio. You have of course to install R first. R is a super famous statistical language. The ggplot2 package is the most famous plotting R package.

When you learn R, I strongly advice you to create different ready to use R scripts or templates of the most commonly used statistical tests and plots. This will make your life a lot easier in the future.

R has a learning curve as I have mentioned that before, so I strongly advice you to learn it before you embark on your research journey. If you know R already from a previous degree, it does not harm to do a refresher on the language before you start your PhD.

If you want to learn how to integrate R plots and tables into LaTeX or use LaTeX inside R plots such as to use LaTeX for captions text, legends etc.., then please check the four blog articles that I have written  on the topic (Articles 1, 23 and 4).

Recommended Books

Bear in mind that these are cookbooks or books that teach R only. These books do not teach you statistics. You need to master Statistics before you read a book on R (See Skill A).

Of course you will definitely find yourself in need of using the ggplot2 R package so a classical and very good book covering the topic is:

For a book that teaches both Statistics and R. It an awesome, enjoyable, funny and easy to read book!:

  • Discovering Statistics Using R by Andy Field.  The author is quite funny and you will enjoy and begin to like statistics if you hated this topic in the past.

Web resources

  1. The R journal which is the official open source reviewed journal for R statistical language
  2. The an awesome website teaching R programming
  3. The R documentation and manuals website
  4. The R Bloggers – R news and many easy to read tutorials on R
  5. Flowing Data – A great website for teaching R, Python and other data science technologies.
  6. Quick R or stats methods website – contains a big number of tutorials covering R and general statistical concepts.

Online video courses

Since we do not live in the stone ages anymore, as I said the fastest way to learn is through a video course. Fastest way does not mean the more thorough way but it can get you started quite quickly.


PS: Don’t buy courses when their prices are more than 10$ (the price of a lunch). As I said before Udemy is a bazaar market for courses, they always have sales so prices of courses will be reduced to 10$ eventually in sales. Plus a lot of instructors can give you vouchers or they usually offer discounts when you buy one of their courses. Bazaar does not necessary mean bad, it means cheap, but yeh! some courses are really bad so avoid spending your money on them. Buy always best seller courses or positively highly rated courses with thousands or hundreds of thousands of students.

Old S and S-Plus Statistical Languages

No one uses S Statistical Language since the open source GNU free implementation/alternative of the S language is nothing but the R language (stated above). But S-Plus is a commercial implementation of S programming language and is used by some researchers in the social sciences.

MATLAB (Cost Money) or GNU Octave (Free)

MATLAB is one of the giants. It is cost good amount of money for sure, so a lot of the folks I know learn MATLAB and its mathematical and plotting programming language either via books, online resources, video or traditional courses,  lectures etc…and then apply the knowledge (same results) in GNU Octave. GNU Octave is a software that provides the MATLAB scientific programming language. The language  is based on “powerful mathematics-oriented syntax with built-in plotting and visualisation tools “. A lot of academic institutions such as Universities have normally MATLAB Licenses but for the folks out there who are disadvantaged (remember this article is for their sake), I advice them to download and install GNU Octave and then use the software as you use MATLAB.

Python Statistical and Plotting Modules

It is a good idea for you to learn the Python programming language even if you are not a Computer Scientist or a programmer. In addition, you do not need to be a data scientist to follow the Python path. Python is an easy language to learn. It is a a “batteries included” language. This section covers the following important Python libraries for data analysis and visualisation: Numpy, Scipy, Pandas, StatsModelsSeaborn and matplotlib. This does not mean that these are the only ones that are normally used. But it is fair to say the aforementioned Python libraries are the most famous and most used by UK PhD candidates. Some add to the list scikit-learn for “statistical learning” a.k.a by the buzz word: “machine learning”.

A side note: I strongly advise you to register on Kaggle, an online community of data scientists and machine learners. Kaggle allows you to access tons of learning material covering Python, Machine learning, deep learning, Pandas, R among others. It hosts also a huge number of data sets ideal for inferential statistics and machine learning. Another cool thing is the  competitions with rewards amounting to thousands of dollars that you can participate in and earn a lot of money.

All the following modules are under the banner of “Data Analysis and Visualisation in Python”. It is better to learn them as part of an overarching course or topic instead of getting bogged down with one library or another. Since, in my experience, you will need to use them in tandem for many  cases. Per instance, you might use in the same Python script, Pandas to import an Excel (.xlsx) file, Numpy to do some scrubbing and manipulation of your data, the stats module in Scipy and Statsmodels for statistics,  seaborn and matplotlib to plot your data etc…

I advise you when you learn these modules to create many ready to use templates to be used later. Per example, a template of how to import data  (when the header row is available or not) from an Excel workbook, another for importing from a CSV (Comma Separated Values) files,  or from a Tab delimited files,  or from a MySQL database etc.  Similarly, create templates for different types of plots and statistical tests. This will make it easier for you to use them later for different research projects.

One last note, for all the Python newbies out there who are not tech savy, there is a cool web tool that will help you greatly learn and execute Python code which is the Jupyter Notebook project. It is used by many academic institutions to teach Python, data science and machine learning courses. It can be installed separately or as part of the Anaconda Data Science platform. Another scientific and analytic computing distribution that is similar to Anaconda is Enthought Canopy.

Jupyter Notebook known previously as IPython notebook is a kind of a shell and a Python interpreter that works on the web i.e. in a web browser. It is a sort of a “computational notebook” that can contain code, executed results including resulting plots, calculations etc.. and can contain markdown text. It should be mentioned that it is not just for the Python programming language, it can connect to kernels allowing you to program in over 40 programming languages such as Julia, PHP, C#, Haskel, Swift, R among others. Jupyter notebooks are mainly used by academics and students to facilitate the learning and the teaching of programming languages and of essential libraries and for testing out small snippets of code.

Google CoLaboratory or Google Colab which is a free Google service (at the time of writing). It is based on Jupyter notebooks that run on Google’s cloud. It is integrated in the Google Drive application. It allows you to use some of the muscles of Google company’s Hardware (GPUs, CPUs, TPUs & RAM) to run data science and machine learning applications written in Python. All the essential Python libraries are already installed. All you need is a web browser.

Numpy and SciPy

Numpy is a package that supplements your data analysis tool set with advanced data structures such as multidimensional arrays and matrices helping you organise and process large amount of data.

Scipy is an eco-system really not one library or a module which contain under its banner other libraries such as Numpy, Pandas and Matplotlib, among others. It is the “scientific python” in itself as its name suggests. It covers whatever you can imagine doing in mathematics, statistics, scientific calculations and engineering.

Per instance, you can use the Scipy stack of tools or eco-system to do a one way ANOVA test via the (f_oneway() method). The stats module of Scipy contains a wealth of useful statistical functions. PhD candidates and researchers also use a famous dedicated library for statistical tests called StatsModels which I will discuss later.


Pandas is a famous Python library for data analysis. The library also provides advanced data structures. It provides ways to import and export data to different files and sources. Pandas-datareader gives Pandas remote access to various online sources such as the World Bank and including financial services such as IEX, Alpha Vintage, Quandl, Yahoo finance among others.

Pandas library also contains some visualisations built on top of matplotlib although it is not the main objective of Pandas to do visualisations. It is better to use matplotlib and seaborn for that purpose.

It is very beneficial to spend some good time on learning well the Pandas and Numpy libraries. Pandas is ideal for importing & exporting data from csv, tab-delimited data, MS Excel, JSON, XML files formats, from databases and online sources etc… It allows you to manipulate & clean data transformed into Pandas data frames which in lay terms look like MS Excel worksheets  if you like. You need also to learn how to create, manipulate, slice, split, stack, iterate over.. numpy arrays. Numpy arrays are used by a big number of
Statistical, Data Science, Machine Learning, and Computer Vision Python libraries. For instance, in OpenCV (a famous Computer Vision library), all the images and video frames are represented  pixel-wise by Numpy arrays .


StatsModels is your Python library for statistics and statistical tests. Name any statistical test and you will probably find it in this module. It covers many essential and commonly used descriptive statistics, statistical tests, and plotting functions. Students usually use it in tandem with the stats module of Scipy.

The following two sections covers briefly 2 awesome and important visualisation Python libraries: matplotlib and seaborn. There are others of course such as Vincent, Plotly, Bokeh etc…  A quick word here: Bokeh & Plotly are interactive open source Python plotting libraries on the web. It might be ideal to use them in scenarios where you want to show data visualization in web browsers to other researchers that you are collaborating with.

Matplotlib and Seaborn are the most stable and most commonly used. Actually the majority of other plotting Python libraries are based on Matplotlib. My advise to you is to learn both Matplotlib and Seaborn. Start with matplotlib library and then move on to learn seaborn. Why? well! matplotlib allows you to draw many types of plots but seaborn have extra types of plots (like faceting plots) and things of that sort that you might find quite handy in your PhD.


Matplotlib is a famous 2D plotting library in Python. The types of plots you can draw with the Matplolib library are just awesome. Please have a look at the examples in this link and you will be astonished. Actually, I regret that I did not use this Python library instead of R from the beginning of my PhD. Now usually you need some knowledge also in Numpy and Pandas Python modules while working with matplotlib so all these 3 modules supplement each others in a way. Per Example, I usually use the  Pandas module to clean up raw data and import it from different sources (Comma Separated Values (.CSV) files, Excel, Tab delimited files vel cetera).

Why matplotlib module is awesome and powerful?

I could write a complete book on this but I just want to mention in this section few advantages that I have found. First, writing my plots scripts in Python as a language compared to R per example is very advantageous. As you know Python is an easy language that makes code extremely readable (one of the main selling points of Python). Second, the concept of matplotlib ‘styles‘ or ‘styles sheets’ is just amazing. matplotlib has styles that make your graphs look like a ggplot R plot (i.e. emulating R ggplot), or like plots of seaborn module, among many other styles. There is a great number of styles that you can choose from. You can also create your own style. So you can specify how the axes should look like, the colours, sizes, labels, the ticks etc.. and then apply the style to your graphs without redoing the work.  Third, matplotlib is amazing in creating grids of plots and subplots (meaning several plots in a grid in one figure), it is  a lot easier to do that than using R. In addition, matplotlib makes it relatively easy to make plots that share axes or to create many axes or dimensions in your graph. Trust me you will need this at some point! Most of all, matplotlib is awesome in creating something called ‘inset plot‘. One cool type of inset plot is called “Zoomed (in/out) Inset plot” (please see next section).

Awesome plots that I was able to make with Matplotlib

There is a graph that I saw once in a academic paper and I was fascinated by it. I was also very frustrated because I wanted also to be able to do something like it for a scenario that I needed. The paper was published in 2003 if my memory does not betray me. I reached a stage where I wanted  to contact the authors to ask them about that. Of course that would have been extremely silly.

How the hell did these folks do that?
How the hell did these folks do that?

Finally an eureka moment arrived and I discovered this type of plot is called “Zoomed-Out Inset Plot“. R has the ability to create inset plots but positioning them is just weird and there is no zooming effect as can be seen in the image above. The only place where I have managed to replicate exactly that plot is with Python Matplotlib. Thus finally I could sleep at night since I knew the secrets of these authors. They used Python Matplotlib. Check this person’s blog , or the Inset Locator Demo2 on Matplotlib documentation, if you would like to know how to create them. If you know other systems that can do that, please leave that in the comments.

On many occasions, you draw a plot on a scale to show the overall trend of the lines but sometimes you have a part of the graph which is either too jammed or too small. The “Zoomed-Out or Zoomed In Inset plots” are ideal for such cases relieving from you the burden of drawing two or more plots.

In matplotlib you can save your plots in different formats as follows:

import matplotlib.pyplot as plt
... code of your plot...
plt.savefig('images/creatingstyles2.png', dpi=600)
#if needed

In the above example, I saved my figure as a PNG, as a PDF and as a PNG while specifying the resolution as Dot Per Inch (dpi). There is always this annoying problem of NOT generating figures worthy of publication quality especially for academic papers, theses and for printed posters. You will always feel frustrated by the quality of the images that you normally generate which normally are ugly and pixelated especially in the case of a big poster, per instance. The DPI parameter might provide some relief for cases where you use your images in MS Word/MS PowerPoint. As a rule of thumb always aim for at least 300 dpi preferably 600 dpi for anything that will be printed such as a poster or a thesis. I found that using graphs or images as PDF provides a better resolution quality images than using PNG or JPEG formats.

If you use LaTeX, there are other better techniques. Please check the article that covers essential LaTeX tips for research students.

Matplotlib Learning Resources

Books, Chapters and Articles




Seaborn is an exciting data visualisation library. Seaborn can let do things similar to what ggplot2 package in R can let you do. The example gallery of some of the plots that Seaborn can draw can be found here. I like the fact that you can add many ‘semantics‘ to any plot you can imagine; sort of dimensions or additional variables to your plot. You can supplement any plot with semantics like “hue” (meaning you add another variable to the plot shown in different colours), “size” ( the variable is shown by size) and “style” (an additional variable or dimension is drawn with different styles of lines or shapes of bins etc…).

Seaborn is based on matplotlib  so you might need also to be familiar with matplotlib. Actually, as I mentioned before, learn data visualisation and analysis in Python as “a whole” not just particular Python libraries or modules. It will then be easier for you to learn, if you need, the specifics of a Python Module.

ggplot2 is very famous in R in doing facet grids or faceting (i.e. plotting many variables both categorical and numerical showing many dimensions and angles). Seaborn allows you to do many variations of that in Python. You can use two parameters (row=X , column=Y) for any plot function under the Sun in Seaborn. These will facet the plot into rows/columns of plots based on the dimensions in X variable or Y variable. This is really awesome! I found that the Python Seaborn module is a lot easier than using the R ggplot2 package.

Gnuplot Tool

This tool is loved by the old folks in my school. This tool is ONLY a plotting tool that creates awesome and very beautiful statistical plots. It works by writing your plot in a specific plotting language. The tool also integrates very well with LaTeX. Actually I should say extremely well.

Gnuplot Power

For a demo of what gnuplot can draw, please have a look at the gnuplot 5 demo plots

To learn how to integrate your Gnuplot graphs into LaTeX documents, kindly check my blog article covering some of the essential tips of LaTeX/BibTeX and BibLaTeX for research students. LaTeX is covered in skill E.

D – Master using bibliography management systems such as Mendeley or Endnote

If you are doing things manually in terms of bibliography inside MS Word: per example writing each reference manually letter by letter, stop it! this is too primitive, too manual, too pre-computers era and a complete waste of time. You need to use a bibliography management software or tool.

Master the usage of at least one bibliography management software from the names mentioned in the next paragraph. Know that all of them have plugins for Microsoft Word and other famous word processors.

There is Mendeleya very good free tool that I used before moving to LaTeX/BibTeX/BibLaTeX, it has neat features and it is good for the MS Word folks out there. There is also Endnote and some Web Browser based tools. There is  Zotero, and RefWorks. There is also Citavi which is also popular. ReadCube is quite used by many people that I know.

I mean you can really choose anything as long as it makes your life easier not harder. Choose what suits you! These tools are created to automate and facilitate the referencing process. Invest learning your software of choice before starting your PhD. The majority of these tools facilitate the insertion of citations in your text. They take care of everything and automate the whole process for you. They allow you to organise your bibliography and to annotate and catalogue the academic papers you have read or the ones you end up including in your reports or thesis. In addition, you can with few clicks  change your bibliography style.

All these tools integrate very well with the majority of word processors via plugins. You need per example to install the Mendeley plugin in MS Word.

Please do not forget to consult at least a referencing guide book even when using these automated tools since a lot of them produce in many instances incomplete or incorrect citations. You can also seek the help of websites such as Cite This for Me and citefast. They would help you to manually fill out the details of a citation and then to export that citation to a format suitable to your bibliography management software such as Endnote or Mendeley.

If you are doing a PhD in the sciences, there is a big probability that you will be obliged to learn LaTeX & BibTeX/BibLaTeX especially here in  the UK because there is a big probability that your supervisor and the academics in your research group will force you to use LaTeX for writing academic papers, reports, presentations, posters, thesis etc. Please ask your prospective supervisor and other PhD candidates in your research group whether knowledge of LaTeX is essential.

LaTeX has a bit of a learning curve so it is strongly advisable to learn it before stepping into a PhD.  In addition, all journals and conferences proceedings require LaTeX and in few instances: I saw some journals give authors only a LaTeX template to work with. Usually, you would have two conference templates one in Microsoft Word and the other is in LaTeX.

E – Master LaTeX, BibTeX and BibLaTeX

NB: You can skip this section, if you do not plan on using LaTeX/BibTeX or BibLaTeX for your PhD. Usually this is the case of a PhD in the humanities and a number of disciplines pertaining to the social sciences. Few instances of PhDs in natural sciences might not need the knowledge of LaTeX. It depends really! If you need to write a lot of complicated mathematical equations, it makes sense to invest in learning LaTeX instead of using Microsoft Word.

Please read my blog article covering the essential tips that all research students need to know when using LaTeX. The article is not a tutorial of LaTeX, nor the material in this section. If you are a newbie to LaTeX, I have suggested some material for you to consult at the end of this section. After you learn LaTeX, it is pertinent to consult the LaTeX article for all tips, and tools that I have found crucial throughout my PhD years especially when it comes to writing papers and my own PhD thesis.

LaTeX pronounced «Lah-tech» or «Lay-tech», is a typesetting system and a typesetting language. According to the LaTeX project’s website, LaTeX is the “de facto standard for the communication and publication of scientific documents”.

BibTeX and BibLaTeX are two very famous references management software and notation languages.  BibLaTeX is more advanced, more fancier than the good old BibTeX. Actually, BibLaTeX is considered the successor of BibTeX. Both in a technical sense are TeX packages that come by default these days with all the major distributions of LaTeX. BibTeX has an executable or a sort of a compiler that is run behind the scenes when you compile your LaTeX document. This compiler is used to parse your BibTeX bibliography notations which represent your sources (papers, books, online material etc…) i.e. entries in the .bib file. Same for BibLaTeX but the executable or the processing program is called Biber. You do not have to worry about any of this as both of them are available by default in major TeX distributions. BibTeX is very old but is still used. BibTeX is great for numerical citation styles such as IEEE, ACM etc… but can be extremely annoying when it comes to using an (author, year) citation style such as (Bakri, 2016) following Harvard, APA etc… or a label-based citation style a.k.a alphanumeric such as [ABC95] (ex: alpha) or a superscript-based citation style such as the one used by the nature journal. When using only BibTeX, you end up being forced to use another LaTeX package such as natbib in tandem with BibTeX. Nowadays, students learn LaTeX then BibTeX  and then move on to learn and adopt BibLaTeX. I believe this is what you should do. BibTeX is still commonly used by the majority of journals & conference proceedings, this is why it is still alive and is still very relevant. Last update of BibTeX was done 9 years ago. So in my opinion you have to learn it first. Don’t be afraid there is not much to learn! You might find yourself obliged to switch later to BibLaTeX which probably will occur at a certain time especially if you will not use simple bracketed numeric citations styles and when you need advanced customisations of your citations.

LaTeX allows you to create very  beautiful conference and journal articles, technical reports, books, theses, dissertations, posters, resumés, leaflets, flyers, business cards, slides presentations (through the Beamer package per example) and the list goes on and on. If you have tried to submit before to any academic workshop, conference or journal, you will realise that they usually provide you with a LaTeX template for the style and layout of the papers in their proceedings. LaTeX is distributed by the servers of CTAN (Comprehensive TeX Archive Network). In there you can actually download many individual packages which are little specialised LaTeX programs that do particular tasks and you can also download their documentation.

Benefits of LaTeX: why using it? Isn’t Microsoft Word just fine?

LaTeX is an excellent tool for producing very good professional looking documents. It a language with some learning curve. The question that you might be asking yourself is: What is wrong with Microsoft Word or other Word processors? My answer is nothing. Microsoft Word is an amazing powerful product. You can write a PhD thesis in it and still get a PhD. If you are doing a PhD in Divinity or Literature or something like that, I believe it is not worth spending any time learning LaTeX so that the document look a bit nicer. MS Word in such cases is just fine.

That has been said, LaTeX has a lot of benefits and powers over Microsoft Word or any other word processor. The following list presents few benefits that make LaTeX very attractive in academia  especially in the scientific fields:

  1. The look and feel of the documents produced by LaTeX is superior to the look and feel that a Word Processor can produce. You might wander why? Well, because LaTeX is like an aristocratic lady. She knows all the etiquette concerning the right amount of spacing, the right and advisable typographical rules used in the industry of publishing, the way captions should be presented. In addition, it knows per example where to place tables or figures (normally LaTeX aims always for the top of the page) among many formatting and style etiquette. There are lot of rules that LaTeX take control of and apply them automatically to your document on your behalf in order to make the document look beautiful and professional. Of course you can overwrite that behaviour but if you leave LaTeX do its thing, in the majority of cases,  you are guaranteed to get a beautiful professional looking article, poster, presentation, academic paper, a book or a PhD thesis.
  2. The biggest of all advantages is the separation of style of the document from the content of the document. This is huge advantage!! LaTeX ensures consistent formatting  and that the style of a document can be changed by simply changing a class file or loading some packages with few lines of code. Imagine doing that with Microsoft Word and imagine you have document of 300 or 400 pages. It will be very tedious to update such document or to change styles. So you would pass by every page and change the styles of headings or whatever. Let me give you a concrete example: You have a thesis of 300 pages. All your figures’ captions are labelled as ‘Figure’. You were asked to change captions to ‘Fig’ but the cross references inside the text should remain ‘Figure’ or “figure” depending on the location in the sentence. What will you do? In MS Word, you would probably do a search all and replace but this might have unexpected and unwanted changes. This effect can be achieved in LaTeX in a single line. That’s it!
  3. LaTeX has advanced typesetting and apparent advantage in typesetting complex mathematical  equations and scientific notations than any word processor.  Imagine you are writing a thesis that require so much Math and symbols. You will find LateX is a lot easier for you than composing complicated equations with the equations’ designers that you find normally in word processors. LaTeX contains even symbols not present and could not be found in any word processor.
  4. LaTeX is the king of cross-referencing chapters, sections, subsections, paragraphs, equations, figures, tables etc… It makes it easy to generate a list of tables, a list of figures, a list of algorithms, a list of equations, a list of programming code listings or list of “whatever you want really”, and update them and customise them in all possible way. It is very easy to create an index, a glossary, a bibliography. You can do back-referencing in LaTeX: a process that I don’t think it is available in MS Word to the best of my knowledge. The point that I am driving home here is that LaTeX contains extremely advanced features that fills out books. In addition, if you use with LaTeX another program which also compile a sort of language called BibTeX and you say you have 400 citations and say you were asked to change the style of these citations. No problem! one line in LaTeX! or asked to reorder them alphabetically: also no problem one line!
  5. LaTeX frankly allows you NOT to piss off reviewers of papers or examiners of your thesis by presenting to them a badly formatted document. Impressions are very important! It is the human nature! LaTeX does not even allow you to have a badly formatted document due to its nature unless you mess with its default settings or you overwrite things that you are not suppose to touch.
  6. LaTeX is the aristocratic lady that knows how to present well what is known as f-ligatures. Ligatures always look beautiful in LaTeX.
  7. LaTeX is not bogged down performance wise by the amount of material in your document. Have you had the experience of your 300 pages MS Word document becoming extremely slow especially when you have a lot of images, tables  and cross references and  to make things even worse a Mendeley plugin that keeps an eye on any changes in the locations of all the hundreds references? That document is bound to crash or at least you have to wait some time until you see changes when you write. In LaTeX you do not have such problem. Actually you can write LaTeX code with the most primitive text editor, you have. All the formatting happens in the output after running the building process of LaTeX.
  8.  A lot of supervisors/advisers especially in the scientific fields demand explicitly from you to use LaTeX. They do not accept anything else. If you are in doubt, ask your prospective supervisor whether he/she wants you to use LaTeX. Some supervisors put that explicitly on their profile pages so that prospective PhD candidates are aware off.  PhDs are very demanding and fast-paced especially in the sciences. You will find yourself writing to your supervisor every week or two a report of some sort.
  9. Certain features necessary for writing PhD theses or Masters’  dissertations are simply not present in MS Word to the best of my knowledge.  A) Do we have short captions and long captions in MS Word? I don’t think so.  I always look at those ugly MS Word theses, where the list of Figures or List of Tables have these gigantic captions that spans lines and lines screaming out. First, if your caption is that long, it means something is probably wrong. Either you need to divide your tables or figures into sub-components or you are putting some sort of a complete legend that should not appear in list of figures or list of tables at the beginning of the thesis. In LaTeX, you can have for each part, chapter, section, subsection, subsubsection, paragraph, subparagraph, figure, table, equation, theorem etc., a short caption and a long caption ( using the LaTeX code: \caption[Short caption goes here….]{Long caption goes here….} ). This means that you can give a short caption that appears only in the List of Figures or the List of tables or the List of theorems, and so forth, and a long caption that appears under the figure or above the table inside the main text. Textual Legends under tables or above figures are very easy to do in LaTeX, you just provide another starred caption at the end of your code before closing the environment: \caption*{Legend can be as huge as you want and guaranteed of not appearing in list of figures or tables}. The star tells LaTeX not to put the caption in the List of Figures or List of Tables. B) Can you do back references in MS Word? I don’t think so. C) The power of glossaries commands in LaTeX like \gls{..}, \acrshort{…}, \acrshortpl{….}, \acrlong{…}, \acrlongpl{….},  \acrfull{…}, \acrfullpl{….}, is not present in MS Word. LaTeX is amazing! imagine you can define glossaries and acronyms with plural and singular forms, with short, long and full forms and with a sorting order. Inside your text, you do not have to remember whether the acronym that is used for the first time is expanded or not; you just write the \gls{…} command and LaTeX makes sure that the first time an acronym is met, it will be expanded automatically.
  10. The prenote and postnote of citations: things like see p.3 or pp. 23-25, cf. (Behr et al., 2001) … it is quite easy to add them and LaTeX knows whether to put a p or pp without you typing it explicitly.
  11. Multiple bibliographies in the same document is extremely easy to do in either BibTeX or BibLaTeX. This is needed for instances such as creating a bibliography or references list for each chapter in a book, or for each article or poem in an anthology etc… or when we want to classify or divide references based on the types of the sources such as books, articles, webography.., or based on a topic or on specific keywords or for specific authors or based on primary vs secondary sources etc… I don’t think that this is easy to do in MS Word or Mendeley or Endnote  as far as I know. PS: In BibTeX on its own, you cannot do multiple bibliographies, you need additional packages such as multibib, chapterbib or splitbib. For BibLaTeX, you do not need any additional packages.
  12. Cleveref ( or clever referencing): makes cross-referencing very easy. It figures out automatically whether you are cross-referencing a figure, an equation, a table, a chapter, an appendix etc. Not to mention, you can change the style, with few lines of LaTeX code so all your “Figure” becomes “Fig.” etc.
  13. … I am sure there are many more…

Hopefully, I have convinced you why you should learn LaTeX and BibTeX/BibLaTeX. Please start learning LaTeX early on ideally before you start since it has a learning curve.

There are disadvantages to be frank with you of using LaTeX/BibTeX/BibLaTeX. The obvious one is the learning curve.

So please make sure you figure out whether you need to learn LaTeX before you embark on a PhD journey. Many supervisors and research groups require such knowledge! Furthermore, in many competitive & fast-paced scientific PhDs especially in the top universities in the UK and the US, there might be simply no time at all to learn LaTeX. Everything is fast and brutal! This is why I wrote this article! So please plan ahead!

Good LaTeX Learning Resources

  1. The Wikibooks LaTeX: awesome articles covering a big range of LaTeX topics although some articles are still not completely written.
  2. ShareLaTeX Documentation:  I like their documentation covering LaTeX.  The documentation is not complete or exhaustive but it is nevertheless easy to read.
  3. Dickimaw LaTeX Books and Articles: Mainly 3 Books: (1) “LaTeX for Complete Novices” , the book covers how to write a PhD thesis in LaTeX and is a marvelous book: (2) “Using LaTeX to Write a PhD Thesis” and “LaTeX for Administrative Work“. The 3 books mentioned are online, free and have PDF versions if you want to print them. They all come with many LaTeX exercises and solutions and all the source code used in the books can be downloaded. I totally recommend them, they did miracles to my understanding of LaTeX when I started learning it.
  4. The Not So Short Introduction to LATEX and the An essential guide to LaTeX usage. Both of these guides are somehow old now but nevertheless contain good material.
  5. CTAN for checking documentations of individual LaTeX packages or individual document classes since books covers only common packages. You can always use the terminal/command line “texdoc” tool to fetch the PDF guide of particular LaTeX package. Type per example, on a your operating system’s terminal or command line per example: “texdoc mathtools” to get the documentation of the mathtools package.
  6. There is a lovely FAQ (Frequently Asked Questions) web site with tons of material on LaTeX and its related tools (BibTeX, makeindex, AMSTeX…) which can be found at texfaq.
  7. I recommend the following books if you need a more in depth understanding of LaTeX: The Latex Companion, Guide to LaTeX (Tools and Techniques for Computer Typesetting), The LateX Web Companion (a somehow old book from 1999 that contains explanations of a lot of cool stuff including how to use LaTeX for Web publishing etc… ) and The LaTeX Graphics Companion (2007).
  8. I will be publishing on Udemy a 21 hours+ introductory course on LaTeX/BibTeX. It will be my first online video course. Hopefully it would be enjoyable and useful for students. I will add the link here later whenever the course is ready and published and I will add coupon discounts here for you to enrol and get a good discount. I was thinking of making this course completely free since I will not be able to have the time to manage it or to support students in terms of answering their questions on a regular basis but then the course took considerable time and time is money.  I have recorded almost half of the lectures now. The course will covers all the basics (Document Structure, Page Layout, text and paragraph formatting, colours, fonts, lists, languages (including How to write in Semitic languages), rotations, tables, figures, floats, captions, headers and footers, footnotes, hyperlink and URLs…), Mathematics, Theorems, notations from some scientific fields, Listings, how to create a glossary, an index, mastering BibTeX, BibLaTeX, bibliography (including hyper-referencing and back referencing), a big number of tools/editors/online editors, a complete section on Tikz, all sort of integration between R and LaTeX (tables & plots) and vice versa, All sort of integration between GNUplot and LaTeX and vice versa, all integration between the Python Matplotlib and LaTeX and vice versa, how to embed 3D models, videos and sound clips in final PDF, Transform LaTeX into HTML and Vice Versa, Transform LaTeX into rich text and MS Word Document and Vice Versa. How to load data from a CSV file or from a datatool (.dbtex) file or from SQL database and how to manipulate the data (sorting etc…). There is a section on Beamer. I will touch a little bit on Macros and how to create packages. I have a complete section on Templates that teaches you how to create a Curriculum Vitae, an invoice, a thesis, a dissertation, a poster etc…. In addition, I have sprinkled couple of quizzes to test your knowledge and there are many exercises we do together.  Realise all this time and I did not even dare to call the course other than an introductory course since I know you can do way more advanced stuff with LaTeX.

F – Master writing good Literature Reviews

You need descent Literature Reviews for many things: for creating research proposals, for publishing academic papers, for your own research and your thesis etc…

Please refer to an article on how to write good critical literature reviews that I have wrote. This article is now a recommended reading by many supervisors & lecturers to a lot of postgraduate students in many universities which was astonishing to me and have put a lot of stress on me to tell you the truth. I was writing for myself and for few of my blog readers, now it is like I am writing for complete classes of students.

Disadvantages and sufferings of the folks outside Academia

The great disadvantage that you might have when being outside academia is that unfortunately you DO NOT HAVE ACCESS to  academic papers published in reputable publishers (as an example for  Computer science: you do not have access to ACM digital library or IEEE or Elsevier or ScienceDirect  or whatever other publisher). Some academics particularly some advisers are hypocrite in the sense that some of them even have the audacity to wonder why it is difficult for someone outside academia to write a literature review in a PhD research proposal? Why are they even wondering? You are sitting on your computer somewhere in Africa, or Asia, you finished your undergraduate or even Masters from a long time ago and you suddenly need to create a PhD proposal and to review a literature which you do not have access to it in the first place? You are not going to buy every article which costs could cost 20$, 70$ each and even more.

One of the most important aims of a PhD research proposal is to propose to further the understanding or to fill out particular gaps in the Literature. The only tool to systematically and scientifically discoverresearch gaps” or what is known as “research void” is through a rigorous and coherent Literature Review (whether Qualitative or Quantitative)[1].

One question for any reader that has an IQ > 80. How the hell are you going to do a rigorous and coherent Literature Review for anything actually not only for PhD research proposals, if you do not have access to the damn digital libraries of your field in the first place!!!???

This applies to all disciplines. Some digital libraries in Computer Science such as the ACM Digital Library;  allow you to pay like 198 USD for a yearly ACM membership plus subscription to the ACM digital library which what you are really after. If you are living in a third world developing country, ACM would give you a good discount. This membership would allow you to have access to all their published papers. Not cheap but not a big deal either especially if you are desperate and especially if you are writing a lot of PhD research proposals and don’t have that access. I respect them for doing this. Other digital libraries of publishers such as Elsevier- Science Direct,  Springer-Verlag, Sage Publisher, Wiley Online Library,  IEEE Xplore, Routledge among many others, have subscriptions with prices per year that are ridiculously high. Actually these subscriptions are only affordable by big academic and research institutions. Not all people who are applying for a PhD are currently doing a degree in such institutions or are staff of such institutions. The absolute majority of universities in 3rd world countries can not afford the pricey subscriptions required by many of these publishers. Not to mention a lot of the readers of this article might be in the industry and want to do a PhD.

Without access to digital research databases of famous publishers, the only thing that you can see in a paper is the title and the 200/300 words’ abstract. Not to mention, due to the enormous hypocrisy found in academia, the majority of abstracts of academic papers are just pathetic sales pitch. I am still talking about reputable journals and conferences not low quality ones. I could mention different examples of such papers but I do not want to be sued by authors. I saw a lot of hypocrisy in academia!  and this is at the top universities. Makes you wonder what is happening at the bottom of the ranking ladder.

Actually, in simple terms, can you buy over 300 papers to do a descent PhD proposal? How unfair is that?  How much hypocrisy do we have in academia? I mean this is why the Open Access movement is extremely important in academic research. I am ready and happy to pay double for conference fees, just to make my academic papers OPEN ACCESS.

This is why the best shot for a student (mainly an overseas student) is EITHER to have someone who can download a considerable set of papers that are relevant for him or her OR to have a willing prospective supervisor to help him/her in writing up the PhD Research Proposal or at least help in acquiring the necessary reading material covering the research topic.

Now there are some universities that offer members of the public access to their libraries. Some offers limited memberships. Many of them can give you a membership card which allows you to borrow a certain number of academic books or printed journals/proceedings. Some university libraries can give you a computer account that allow you to access few online journals. The scope of what you can access is very limited. Now such computers are connected to the University internal network and thus would give you by consequence access to journals at least limited access. Again this option is not widely available in all countries or for all the universities. Here in UK, I saw only very few universities that do that. I greatly admire and respect these universities! In UK, you might need what is know as a National Entitlement Card to obtain a library membership in such universities.

Side Advice: If you are coming from a 3rd world country, it is better for you, to do a Masters first in the University that you are targeting for a PhD. If you manage to do that, make sure you take advantage of your Masters to acquire the skills needed to undertake a PhD and make sure and that you meet and know all the staff of your school in other words do not isolate yourself. In this way you can choose with deliberate thinking, a stable and descent supervisor/adviser that you know with certitude that you can definitely do a PhD with him or her. Make sure you both get along very well. You will work with him/her at least 3 to 5 years. Make sure also that he/she has research funds and he/she can help you in applying for a scholarship.

Please have a look at Skill A in this article which many consider it as the most important among all the skills, because unfortunately in the British system, supervisors are given sometimes powers that should not be given to them. There is this centrality of the role of the supervisor in the system. If things does not work out maybe because of the supervisor behaviour or wrong doing not because of you necessarily, things might go south for you. Every place on the planet except UK and few other countries, call these academic staff  “advisers” not “supervisors” and this is done for a self-evident reason which is to emphasise what their main job or duty is. There is a big difference in semantics between being an “adviser” for someone and being a “supervisor” on someone. After all in UK PhDs , we already have reviewers who have as main duties to objectively  judge students’ performance and judge the progress of their research, in addition, in the British system, reviewers must not be supervisors or previous supervisors of the PhD candidate. Obviously the same applies to the candidate’s PhD thesis and Viva Voce examiners. This is in the aim of achieving complete transparency and objectivity. Add on that, reviews for a PhD candidate occur every year once or twice. So in my opinion, the role of “supervisor” (in its current British semantics) is a extremely stupid and redundant role given to the wrong person who his/her job should be on your side i.e. an advisory job like that of a consultant who has a vast research experience that you do not have.

Some bad supervisors act as if the PhD candidates are their own employees and that is strongly against universities policies especially if the doctorate is not an industrial one. We should lobby so that this terminology is completely removed from the UK academic system, because it is misleading.

G – Master Research Methods and Methodologies

I strongly advise you to read from cover to cover at least one book on research philosophies, methodologies, methods and instruments in your specific field of research before you start your PhD.

You need to know how to formulate your research questions, and hypotheses. You need also to know how to plan and design your research on a solid ground. What methodology and methods will you espouse and why? You need to defend the choice of your research methods and in similar vein, you need to be aware of the criticism that each method has since your PhD will be assessed on the integrity and validity of your adopted research methodology. Research methods books are very divers. They are also very essential since they explain and expose the pros and cons from the literature of each research method in your particular field of research. These books usually contain few chapters to help you write a research proposal although there are dedicated books for that particular task.

Some fields have their own research methods books. Per instance, if you will be doing a PhD in Human Computer Interaction (a sub-field of Computer Science), there is a great book titled: Research Methods in Human-Computer Interaction.  It is not a big book. I have read this book from cover to cover and I found it well written and very easy to understand.

There are many research methods’ books for fields of psychology, psychiatry, economics, management, finance, business, medicine, architecture etc.. even many on sub-fields of disciplines. Some books are specialised in only one particular research method such as Grounded Theory or ethnography etc.. across many disciplines. That has been said you will also find many general high-level books on the topic of research methods/methodologies that are not specific to any field or discipline. I usually find them too vague, boring and contain a lot of fluff so I strongly advice you to stick with the ones pertaining to your particular research topic or field.

Before you start doing any research, I strongly advice you to read at least one research methods book as I mentioned before. The first thing you should do to tell you the truth! You should know the difference between  “research methodology“, “research method“, “research paradigm” and “research instrument/tool“. You should also know very well the different research methods in your field of research. Actually, the validity of anything you find is based on the soundness of the methodology, the methods and the instruments that you have employed in order to come up with your results and Ph.D examiners love spending a lot of time hammering you with tough methodology/methods questions!!!! My supervisor was examining a PhD thesis in a Viva Voce and he told me he spent 45 minutes just asking the candidate questions about the methodology chapter, so be aware!!!

Few examples of research philosophies: The positivist philosophy, The interpretive philosophy, the phenomenology philosophy, the participatory philosophy etc…

Few qualitative and quantitative methodologies could be: experimental/empirical, survey, developmental, action research, ethnography, critical ethnography, case studies, evaluations, historical research, policy research etc…

Many candidates even at a PhD level, do not even know what is the difference between using inductive research approach, deductive research approach or abductive research approach (Deduction/Induction/Abduction). A lot of people have never heard even of the word “abduction” or abductive reasoning/abductive research. If you do not know, do not worry, have a look at these lovely YouTube Videos (Video 1, Video 2Video 3, Video 4, Video 5) that explain the difference very well.

You will definitely be asked a lot of questions concerning methodology in your Viva Voce oral examination. Per example, suppose you have collected a lot of data. You have also studied the patterns and have formulated a theory that tries to explain the patterns in your data. Someone might ask you: what did you use, an inductive or an abductive approach?  what will you reply? Abduction is a form of induction actually the form that represent “Inference to the Best Explanation“. So which one did you employ? Have you thought of that?

Sampling questions tend to appear quite often in Viva Voce examinations. What type of sampling (probability/Non-probability) have you used and why? It is important to have a cogent rationale (the why?) of everything you do in your research. There are some software that allow you to calculate the minimum sample size required for different statistical analyses. Questions of validity and reliability are also quite common in Vivas: can your research be replicable? A research methods book in your particular field will tackle and explain all this so reading one or more books before you start your PhD is of the utmost importance.

Please refer to the thesis assessment and Viva Voce preparation article for a lot of questions that were asked in UK Viva Voce examinations and for tons of advice. The article is quite rich and I believe it even merits being read before a Ph.D so you can prepare very well and know what to expect. I assure you that you will find many of the questions asked in Vivas cover methodological research approaches.

In addition, you can enroll in courses that teach research methods in your field of research. Many schools and universities around the world offer such courses at an undergraduate degree level and at a postgraduate degree level (Masters and PhDs) so please take advantage of them. Similarly, I know also that unfortunately a lot of universities and graduate schools around of the world don’t offer them and students are not even aware of the research methods in their fields. The problem is that in the American credits system, which is used by the majority of universities around the world, a lot of people can opt out of taking in my opinion essential courses such as a research methods course and still graduate. Even here in the UK, for many disciplines, curricula do not cover research methods not in the sub-honours level, nor in the honours level nor even in the Masters level which is weird and stupid. Imagine!

In addition, you can enroll in workshops or even online video courses covering research methods on many platforms just to mention few examples such as Coursera (please check this list) or Udemy (please check this list). Although I believe at the time of writing, the current available online video courses on the topic are very limited and in many cases even very primitive so reading books and published peer-reviewed papers is the only way to go for the unfortunate lay person. I have been told by few colleagues about epigeum but I am not sure how much use they can be, especially if your institution is not subscribed to their programmes.

The following sections present few suggestions of books and online resources on research methodologies and methods that are cross-disciplines and research methodologies/methods specific to certain fields:

Good Books that are cross disciplines

You can of course buy one of the following general high-level books  but I believe if you are researching a specific field, it is better for you to buy or borrow from the library a book that covers specifically your research field instead of reading a book that is too general when it comes to covering research methods. Ask your supervisor/adviser, or other academics or PhD candidates or cognoscenti in the field to suggest a good research methodologies/methods books and essential academic papers to read:

  1. Research methodology: Methods and techniques by Kothari
  2. Research methodology: A step-by-step guide for beginners by Kumar [Covers health, education, psychology, social work, nursing, public health, library studies and marketing research]
  3. How to research by Loraine Blaxter
  4. Introduction to Research – Understanding and Applying multiple strategies by Elizabeth Depoy
  5. The Good Research Guide – For small-scale social research projects by Martyn Denscombe (6th Edition) – a very good book for the social sciences folks
  6. Quantitative Methods in Social Science by Stephen Gorard [cross discipline book but only for social sciences such as economics, sociology etc.]
  7. Keywords in Qualitative Methods by Michael Bloor et al.
  8. Your Research Project by Nicholas Walliman
  9. Case Study Research and Applications: Design and Methods by Robert Yin

Online Video courses/YouTube Channels

There is literally thousands of research methods books specialised in thousands of disciplines and fields of research. The following is a very small list of books pertaining to very few disciplines (Computer Science, Information Technology, Management, Finance, Accounting, Psychology/Psychiatry, Medicine/Health etc…). Please ask for suggestions of good discipline-specific research methods books to read before you start your Ph.D from previous Ph.D candidates, your supervisors/advisers and other researchers.

Computer Science Methodologies/Methods Books

There is a scarcity of books that cover research methodologies and methods in Computer Science as a whole i.e. books that play the role of an encyclopedia of research methods or a textbook that “rule them all” if we can say that. Nevertheless, there are few awesome people who wrote books on research methodologies/methods for their respective sub-fields of Computer Science.

Human Computer Interaction (HCI)
  1. Research Methods in Human Computer Interactions by Jonathon Lazar et al. – I read this book cover to cover and I find it quite informative! this is book is recommended by the majority of HCI supervisors. A must read if you will be doing research in HCI. It also helps a lot folks who are investigating other user-based topics in Computer Science such as Quality of Experience (QoE), Usability/Accessibility and other research topics which are similar in nature.
  2. Research methods for human-computer interaction by Paul Cairns and Anna L. Cox
  3. Doing Better Statistics in Human-Computer Interaction by Paul Cairns
  4. Universal methods of Design – 100 Ways to Research Complex Problems, Develop Innovative Ideas, and Design Effective Solutions by Bella Martin and Bruce Hanington – to tell you the truth this book is for both industry and academia with an emphasis more on industry. The 100 methods explained here are used in a variety of user design/user testing scenarios. I read this book cover to cover and I learned a lot. It is not really a research oriented methods book in an academic PhD kind of sense such as Lazar et al. book but despite that it will provide you with several research methods used in industry that might or might not be applicable to your research. NB: Each method is explained very briefly.
Artificial Intelligence
  1. Empirical Methods for Artificial Intelligence by Paul Cohen
System Performance, Simulation and Quality of Service Methods

Many supervisors in UK universities recommend Raj book to their students to read. If you are doing a PhD in Computer Networks’ Quality of Service (QoS) or any system performance studies, the following list is for you:

  1. The Art of Computer Systems Performance Analysis: Techniques for Experimental Design, Measurement, Simulation, and Modeling by Raj Jain – This book is the holy grail on the topic.
  2. Measuring Computer Performance: A Practitioner’s Guide by Lilja
  3. Systems Performance: Enterprise and the Cloud by Brendan Gregg
  4. Performance Modeling and Design of Computer Systems: Queueing Theory in Action by Mor Harchol-Balter
Some academic papers on Research Methods in Computer Science
  • Saunders, M. N. K., and P. C. Tosey. “The layers of research design.” Rapport Winter (2013): 58-59 -This paper was suggested as a reading by my supervisor, it is not CS related but it contains useful info covering a lot of fields.
  • Demeyer, Serge. “Research methods in computer science.” ICSM. 2011.
  • Dodig-Crnkovic, Gordana. “Scientific methods in computer science.” Proceedings of the Conference for the Promotion of Research in IT at New Universities and at University Colleges in Sweden, Skövde, Suecia. 2002.
  • Perry, Dewayne E., Adam A. Porter, and Lawrence G. Votta. “Empirical studies of software engineering: a roadmap.” Proceedings of the conference on The future of Software engineering. ACM, 2000.
  • Easterbrook, Steve, et al. “Selecting empirical methods for software engineering research.” Guide to advanced empirical software engineering. Springer, London, 2008. 285-311.
  • Höfer, Andreas, and Walter F. Tichy. “Status of empirical research in software engineering.” Empirical Software Engineering Issues. Critical Assessment and Future Directions. Springer, Berlin, Heidelberg, 2007. 10-19.
  • Amaral, José Nelson. “About computing science research methodology.” (2011).
  • Holz, Hilary J., et al. “Research methods in computing: what are they, and how should we teach them?.” ACM SIGCSE Bulletin. Vol. 38. No. 4. ACM, 2006.
  • Hassani, Hossein. “Research methods in computer science: The challenges and issues.” arXiv preprint arXiv:1703.04080(2017).
  • Shaw, Mary. “What makes good research in software engineering?.” International Journal on Software Tools for Technology Transfer 4.1 (2002): 1-7.
  •  Zelkowitz, Marvin V., and Dolores R. Wallace. “Experimental models for validating technology.” Computer 31.5 (1998): 23-31.
  • Runeson, Per, and Martin Höst. “Guidelines for conducting and reporting case study research in software engineering.” Empirical software engineering 14.2 (2009): 131.
  • Wegner, Peter. “Research paradigms in computer science.Proceedings of the 2nd international Conference on Software Engineering. IEEE Computer Society Press, 1976.

Information Technology (Management & Technology)

Three volumes that are very detailed set of academic papers from the Journal of information Technology (JIT)  – just waiting for a good Samaritan to digest, compact them and put them in a textbook for students.

  1. Enacting Research Methods in Information Systems – Volume 1  (covers Critical Research, Grounded Theory, and Historical Approaches)
  2. Enacting Research Methods in Information Systems – Volume 2  (covers Interpretive Approaches and also explores Action Research)
  3. Enacting Research Methods in Information Systems – Volume 3  (covers Design Science Approaches and discusses Alternative Approaches including Semiotics Research, Complexity Theory and Gender in IS Research)

Business, Management and Marketing Research Methodologies/Methods’ Books & online resources

The best and the greatest book on the topic which is the classic of the classics, is the book of Bryan & Bell – a must read for the folks in Business, management and even economics. The following list elucidates few recommended books for Business research Methods:

  1. Business Research Methods by Emma Bell, Alan Bryman and Bill Harley – in the newer edition Bill Harley wrote few chapters – if you are in a hurry you can buy the old 2011 version.
  2. Research Methods for Business Students by Saunders et al. is a GREAT book on research methods. If half of the supervisors in UK recommend Bell & Bryman book, the other half recommend this one.
  3. Mathematics for Economics and Business by Ian Jacques – Now this book, according to a friend of mine who is doing a PhD in Economics, was strongly suggested for reading by her supervisor together with Bell and Bryman book.
  4. A dictionary of business research methods by John Duignan (kindle version and paid online version)
  5. Research methods for managers by Gill and Johnson
  6. Handbook of qualitative research methods in marketing by Russell W. Belk.

Online Video Courses/YouTube Channels

  • Research Methods for Business Students Udemy course – is somehow of average quality. The literature review module of the course is too primitive and even wrong. The course is only a small taster video course when it comes to PhD candidates and is not authoritative in anyway as the instructor clearly states it is for a Masters and Undergraduates. I think it is also a taster for Masters. The course is literally based on Saunders et al. book (Research Methods for Business Students), it misses many things including research philosophical stances. You still need to read the lovely Saunders et al. book or one of the other classical textbooks that I have suggested previously.
  • YoutTube channel called Meanthat has a playlist on research methods for business.

Finance, Economics and Accounting

  1. Research Methods and Methodology in Finance and Accounting by Bob Ryan
  2. Research Methods in Accounting by by Malcolm Smith
  3. Handbook of Research Methods and Applications in Empirical Finance by Bell et al.
  4. Research Methodology in Applied Economics: Organizing, Planning, and Conducting Economic Research by Ethridge
  5. Mathematics for Economics and Business by Ian Jacques – Now this book, according to a friend of mine who is doing a PhD in Economics, was strongly suggested for reading by her supervisor together with Bell and Bryman book.
  6. Quantitative methods for Finance by Keith Parramore and Terry Watsham
  7. Quantitative Methods: for Business, Management and Finance by Louise Swift and Sally Piff
  8. Analysing Quantitative Data for Business and Management Students by Charles Scherbaum and Kristen Shockley

Human Resource Management

  1. Research methods in human resource management by Valerie Anderson

Psychology/Psychiatry Research Methodologies/Methods Books

The lovely psychology discipline has a plethora of books on research methodologies and research methods. Ask your supervisor or any cognoscente in the field to suggest a book. Some books on research methods in the field of psychology are stated below. Please if you have any suggestions, kindly write them in the comments:

  1. Research Methods and Statistics in Psychology by by Hugh Coolican – Buy the new edition (2018) – it is amazing!
  2. Research Methods and Statistics in Psychology by by Haslam and McGarty
  3. Research Methods in Psychology by David Elmes, Barry Kantowitz, Henry Roediger III
  4. Statistical Methods for Psychology by David C. Howell
  5. Statistics and Research Methods in Psychology with Excel by Verma
  6. Research Methods in Psychology: Investigating Human Behavior by Paul G. Nestor and Russell K. Schutt
  7. Research Methods in Psychology: Evaluating a World of Information by Beth Morling.
  8. Research methods and data analysis for psychology by  Stuart Wilson and Rory MacLean
  9. Discovering research methods in psychology : a student’s guide by Sanders
  10. Analysing Qualitative Data in Psychology by Evanthia Lyons & Adrian Coyle

For Psychiatry:

  1. Research Methods in Psychiatry  by Chris Freeman and Peter Tyrer

Information Studies, Knowledge/records Management and Library science

  1. The Facet LIS Textbook Collection: Research Methods in Information by Alison Pickard – this book is great. I learned a lot from it. It has very easy explanation of the research philosophies. It has a great chapter on reviewing the literature which contains good advice. This book teaches you how to choose your sample, teaches you statistical tests among many other goodies.
  2. Research Methods in Library and Information Science by Connaway and Radford.
  3. Research methods in information by Alison Jane Pickard

Political Science

  1. Political Science Research Methods by by Johnson and Reynolds
  2. Political Research: Methods and Practical Skills by by Sandra Halperin and Oliver Heath
  3. Research Methods in Politics by Burnham et al.

Geography/Human Geography

  1. Research Methods Geography: A critical introduction by Gomez and Jones
  2. An Introduction to Scientific Research Methods in Geography and Environmental Studies by Daniel Montello and Paul Sutton
  3. Qualitative research methods in human geography by Iain Hay


  1. Social Research Methods by Bryman
  2. A dictionary of social research methods by Mark Elliot et al. (kindle version and paid online version)


  1. Research methods in anthropology: qualitative and quantitative approaches by H. Russell Bernard


  1. Research Methods in Human Rights by Bard A. Andreassen


  1. Research Methods in Education by Cohen et al.
  2. Research Methods and Methodologies in Education by Coe et al.
  3. Introduction to Research Methods in Education by Punch and Oancea
  4. Research methods in language and education by Kendall A. King
  5. Visual research methods in educational research by Julianne Moss and Barbara Pini
  6. Research methods for applied language studies by Richards et al.
  7. Research methods for English studies by Gabriele Griffin

Theology and religions

  1. The Routledge handbook of research methods in the study of religion by Michael Stausberg and Steven Engler

Encompassing many social sciences fields/Too Specific

  1. Research Methods in Indigenous Contexts by Arnold Groh.
  2. Research Methods in the Study of Substance Abuse by VanGeest et al.
  3. The SAGE handbook of social media research methods by Sloan and Luke
  4. Handbook of qualitative research methods on human resource management : innovative techniques by Keith Townsend et al.
  5. Research methods for creating and curating data in the digital humanities by Matt Hayler and Gabriele Griffin
  6. Research methods for reading digital data in the digital humanities by Gabriele Griffin and Matt Hayler

Health, Medicine & Nutrition

  1. Introduction to health research methods  by Jacobsen
  2. Research methods in health: investigating health and health services by Ann Bowling
  3. Clinical research methods for surgeons by David F. Penson and John Wei
  4. Handbook of health research methods : investigation, measurement and analysis by Ann Bowling and Shah Ebrahim
  5. Nutrition Research Methodologies by Lovegrove et al.


  1. Pharmacy practice research methods by Zaheer-Ud-Din Babar


  1. Research Methods for Architecture by Raymond Lucas
  2. Architectural Research Methods by Linda N. Groat

Few examples of books for other fields

  1. Research Methods for Cognitive Neuroscience by by Aaron Newman
  2. Research Methodologies and Practical Applications of Chemistry by Pogliani et al.
  3. Research Methods and Applications in Chemical and Biological Engineering by Pourhashemi et al. – This book is not released yet and I found it in an online search only.
  4. Engineering Research Methodology: A Practical Insight for Researchers by Deb et al. – This book is not released yet and I found it in an online search only.

Books that cover in detail only one research method (across many disciplines)

These types of books are targeted to a specific research method across many disciplines, in other words, not really tied to a particular field. The following is just a example to show what they could be about. Please search yourself for what you need or might need.

Grounded Theory

The bible on grounded theory is that of Strauss and Corbin:

  1. Grounded Theory in Practice – an amazing book and a legendary classic suggested as a reading by many supervisors.

There are many books tailoring specific research fields. The above suggestions were just few droplets from the ocean. Most importantly from all this, is that you MUST read at least one book like these before you start your PhD. It is very crucial. Not knowing your research methods in your specific field is extremely unwise.

H – Master analysing qualitative data & the usage of a Computer Assisted Qualitative Data AnalysiS (CAQDAS) Software

NB: This applies only for the folks who will be doing qualitative research or mixed methods research.

You might end up collecting and analysing qualitative data such as interviews material, participant observation etc.. If this is the case, you need to master a CAQDAS software to help you in analysing such data.

There are of course many considerations that you should be aware of depending on the qualitative analytical technique that you are adopting whether it is thematic analysis or template analysis or grounded theory etc…

Many CADDAS systems also analyse mixed methods data: the category that involves both quantitative and qualitative data. An example of such CAQDAS is MAXQDA which is in my opinion the most advanced software that I have seen and worked with in this category.

CAQDAS are powerful and enjoyable to use. They can import data from PDFs, word processors and textual documents of all types, Microsoft Excel, Comma Separated Values (CSV), surveys results (including from services such as SurveyMonkeyQualtrics…), from SPSS, pictures, e-mails, memos, audio recordings, videos, web articles or wikis, bibliographic metadata of citations from Mendeley, Refworks, Zotero, Citavi, Endnote, BibTeX etc.., Notes from software such as Evernote, Microsoft OneNote etc. They can scrape, import and code textual data from social media web sites such as Twitter (importing tweets, #hastags, @authors etc.), YouTube (i.e video comments & video transcripts) and Facebook. Data imported to these tools could be interview transcripts, academic papers, survey results, audio recordings or their transcripts etc …

These tools do very powerful analysis on your qualitative data in whatever form you want it and then visualise the data as mind maps, project maps, concept maps, word clouds (a.k.a tag clouds), cluster analysis graphs among many other graphs that aggregate/summarise qualitative data and show them in a meaningful manner.

That has been said, these software do not do that automatically like magic! It is NOT like you throw your interview textual data in a CAQDAS and it would magically tell you all the patterns that can be discovered! We did not reach that level yet! Actually if this is the case you should be concerned since you will not find a job as a researcher!

You have to code your qualitative dataCoding in qualitative research has nothing to do with programming languages.  Coding actually is literally a whole science and there are many books covering the topic. 

In the simplest terms possible, coding in qualitative research is classifying and categorising textual qualitative data by using codes in the sense of semantic labels to define pieces of text. Coding is also a term used in quantitative research and designates using a number understood by the researcher instead of a textual categorical variable (per example: Male gender = 1 and female gender =2) and this for the purpose of statistical analysis.

Do not forget when you adopt a coding framework to make sure that you can assess its transparency and whether it can be applied by others i.e. make sure to assess its inter-rater reliability. This is commonly done by measuring the Cohen’s Kappa. Have a look at these informative YouTube videos (1, 2, 3) covering Cohen’s Kappa and these videos for calculating it using Excel and using SPSS.

As a side note: if you want to measure the internal consistency or reliability of “coding” quantitative data (Example: coding in a questionnaire’s context), you need to learn how to calculate the Cronbach’s alpha coefficient. To learn how to do that in this particular context (i.e questionnaires) please see this YouTube video using Excel and  YouTube video 1 & video 2 using SPSS. It is an easy process.

Coding qualitative data is facilitated by CAQDAS systems such as NVivo, MAXQDA, ATLAS.ti, QUALRUS among others.

Per example, suppose you have an interview text imported into Nvivo, you would pass through the text and code parts of it under semantic categories. You can nest codes and have parent-child relationships between codes. You need sometimes to code every line or two of your textual data in grounded theory research per example. Now the beauty is not in this task, actually it is daunting task. The beauty is seen afterwards because CAQDAS systems can give many powerful analysis tools both visual and non visual to study patterns and interrelationships. In addition, they provide you with a complex code-based querying system.

CAQDAS tools

Some CAQDAS systems are free and others are proprietary and cost money. The following is list of the most common tools following the suggestion of a friend  of mine doing a PhD that requires their usage. It is advisable to take the opinion of your supervisor/adviser on what software package you should be familiar with from the start of your work or even before your PhD starts and it is even advisable to learn the software first and then start to collect you qualitative data since many of these tools dictates how data should be used, stored, coded and analysed.

QSR Nvivo

NVivo is a very famous software that contains many powerful tools for the analysis of qualitative data. It is recommend by many supervisors/advisers. The cheapest student license per year cost 60 quid (so it is not cheap!). Many universities have subscriptions for this tool since it is quite powerful and famous. The following are few resources that could help get started with NVivo.


Online video courses/YouTube channels


MAXQDA is super super super powerful. I have used it myself and used NVivo in a course aiming to familiarise myself with the myriad CAQDAS systems out there. I can tell you right away between the two, I would go with MAXQDA. NVivo is famous and this is why a lot of folks recommend it but I found personally MAXQDA 2018 is a lot more powerful than NVivo 12. Before you decide please take the advise of your supervisor/advisor and other PhD candidates and other researchers who did similar qualitative or mixed methods research as you will be doing in your PhD. You can have a look at what the software can do for you: here. MAXQDA allows you to analyse qualitative and mixed methods data.

Online Video Course/ YouTube Channels

  1. Qualitative Data Analysis using MAXQDA Analytics: Hands On –  is a Udemy online video course that is quite good. The accent of the instructor is a little bit hard to understand and you will feel that he is wasting time in the beginning of the course explaining the features of the software but despite that the course is rich. This is an absolute beginner course in my opinion, it barely scratch the surface of what you can do in MAXQDA.
  2. MAXQDA VERBI YouTube channel contains many video guides created by the MAXQDA team that teaches you how to use MAXQDA. Please have a look at it. You will be amazed by what you can do with MAXQDA.


ATALAS.ti is another famous CAQDAS that helps you analyse qualitative data. It allows you to code, and annotate your textual data and to visualise relationships between them.

Video guides


There are other CAQDAS systems out there. I can give you as honourable mentions: WordStat and QUALRUS. Check them out if you are interested.

General resources for coding, Qualitative & Mixed methods research

Please leave a comment if you have any useful suggestions that are germane to this section.

Books to read


  1. The Coding Manual for Qualitative Researchers by Johnny Saldana – this book was recommended by a lot people who I have asked concerning this topic – a must read before you start coding your qualitative data.
  2. A Step-by-Step Guide to Qualitative Data Coding by Philip Adu

Analysis of qualitative data

  1. Using Software in Qualitative Research by Silver and Lewins
  2. Qualitative Data Analysis: A Methods Sourcebook by Miles et al.
  3. Qualitative Content Analysis in Practice by Margrit Schreier
  4. Handling Qualitative Data by Lyn Richards

Online Video Courses on Qualitative Research

  1. How to analyse Qualitative data  – a Udemy course created and presented by Dr Jaroslaw Kriukow who was involved in many research projects in both the University of Oxford and University of Edinburgh.
  2. Introduction to Qualitative Research Methods

I – Master critical academic writing/critical thinking

This is an essential skill to have for any type of PhD or for any academic degree for that matter. You have to master academic writing.

You have also to learn and to acquire the critical thinking skill. Critical thinking is a skill that makes you a better person in life. It makes you a rational thinker.  It is a skill that can be learned and honed with time. This is not poetry here! This skill is crucial if you want to survive in any hypocritical environment. Some academics will throw at you all sort of logical fallacies and if you are not equipped, you might be fooled quite easily. As discussed in Zobel book [5], strawman and red herring fallacies are common in research and among academics.

Learn how to detect logical fallacies in research papers or in conversations and how to avoid them yourself in your writing and in your day to day life. You have no idea of how much logical fallacies I detect on a daily basis!. Some people find that quite charming about me while others find it extremely annoying and avoid saying anything in front of me.

You need to be able to detect fallacies committed by writers in articles and books. In the scientific disciplines, there are many fallacies committed, most common are the fallacies pertaining to the statistical fields: small sample or biased sample, rushed generalisations, over generalisations etc. There is also a common type of fallacies dubbed as “perceptual fallacies” when it comes to using intuition that “feels right” but is not necessarily right especially in the fields of statistics and probability. Confirmation bias is also very common where many researchers look only for the evidence and for data that confirm their hypotheses or points of view and avoid considering any contrary evidence.

Please learn logical fallacies! It is crucial for your success as an academic, as a PhD candidate, as a writer and as a presenter. There are many cases where I have discovered that even my supervisor was pulling on me a red herring to avoid focusing on a specific argument since he can not win it. Red herring and Tu quoque is used commonly by many politicians. It is very enjoyable to spot them in political speeches.

You need to be able to have full proof arguments in your PhD thesis. Remember a PhD is a Doctor of Philosophy in X – X being the field. You should not commit any logical fallacies yourself. You should become a human detector of logical fallacies.

Master the detection and avoidance of Logical Fallacies

This is an extremely important skill as I mentioned before. The following are resources, books and online courses that I found myself very enjoyable and which taught me a lot when it comes to this topic.

Recommended Online courses on Logical Fallacies

Recommended YouTube Channels

  • Philosophy Tube –  is an amazing YouTube channel about philosophical debates. It also contains a section for discussing many logical fallacies committed in argumentation.
  • Philosophy Vibe – is a beautiful cartoonized channel about philosophy. It contains many videos on logical fallacies
  • Teach Philosophy – contains many logical and critical thinking videos.
  • Wireless Philosophy – also an amazing channel.
  • YouTube Channel – Ah! Waw! this channel is tremendous when it comes to squeezing you mind. You can learn also logical paradoxes.
  • Nerdwriter1 channel – is a great channel for people who want to learn effective writing particularly essay writing.

Recommended websites

Recommended books

Master Rhetorical devices

Rhetorical devices are like a knife you can use them for good or for evil 🙂 They are the techniques used in persuasion by writers, politicians, advertisement industries etc. Learn how to use them in your writing for persuading others of your arguments. Use them in moderation and do not rely on them. A clever academic will detect quickly if the intention of the usage is to mislead. Nevertheless the best pieces of written work in history are works that have used rhetorical devices. These devices aim to make your writing eloquent and enjoyable.

Recommended resources

  • The Art of Communicating Eloquently – a great Udemy course on rhetorical devices which gives you tons of examples from actual speeches, writings etc… The accent of the instructor is a little bit difficult to understand but nevertheless it is a very good course.

You need to learn how to write well precisely how to write “academically well” if we can say that. I have written  a great article about writing skills that you need for your PhD, or for any academic degree. Some of the advice in the writing skills article are present here.

Learn the academic language of your field/ Learn academic writing

Knowing normal English is not sufficient. Recall one of the requirements for assessing a PhD thesis is that the english of the thesis has to be of a scholarly academic quality. Academic writing is a skill that you can actually learn and that you can hone with time and with a lot of practice. No one was born with it and it is NOT a gift. It is a learned skill.

How to learn writing in a scholarly academic language?

Well, the most efficient method to learn academic language comes from reading tons of published peer-reviewed academic papers in your topic or your field. You will learn a lot from scholars in your particular field. You will learn what “terms” and “phrases” to use when describing your methodology, your results and your conclusions. You learn how literature reviews are done among many other things. If you check my article on how to write literature reviews, you will find that I have dedicated in my method, a field for academic english expressions gathered from the papers or books that you read.

To drive this point home, what I usually do when I am reading a paper is to write down what research methodology/ research method(s) that  the authors have employed. I record the main research outcomes, the limitations etc… In addition, I usually capture the academic language ie the academic english expressions used in the paper the way a phrasebank is designed. So I would learn that instead of saying: ” In this experiment, the X research method was used“. The term “used” here is too informal for an academic setting. You can say more elegantly:  “The X research method was employed or was espoused”. Hope you get my point!

The following are some of the very enjoyable books & online resources that I have read either in part or in full when I was doing my PhD and I have found them very informative:

  1. Academic Writing for Graduate Students: Essential Skills and Tasks by John Swales & Christine Feak
  2. Academic Writing and Publishing: A Practical Handbook by James Hartley
  3. Handbook of Academic Writing by The Rowena Murray & Sarah Moore
  4. Writing a Graduate Thesis or Dissertation by Lorrie Blair – a small and very good book.
  5. The Quick Fix Guide to Academic Writing by Phillip C. Shon
  6. Scientific Style and Format by the Council of Science Editors
  7. Success in Academic Writing by Trevor Day

Academic Phrasebanks

There is a cool website that you should definitely bookmark and always use especially if you are not a native English speaker. Well even if you are!. The University of Manchester Academic Phrasebank, website which contains all those needed academic phrases and expressions that authors use for introducing work, reviewing the literature, describing methods and methodology, reporting results, discussing findings and writing abstracts and conclusions. In addition, the website contains tons of academic phrases to contrast ideas, to express being cautious or critical, to classify/contrast and to list etc… mainly these phrases are compiled and taken from around 100 postgraduate dissertations/theses completed at the University of Manchester, and from hundreds of academic papers. There is an equivalent PDF booklet that you can purchase.

In addition, the booklet contains tips on academic style, grammar and sentence structure but unfortunately it costs money unless if you are a staff or student of University of Manchester. Nevertheless, I totally recommend the website and it is worth even to buy the PDF which costs around 5£, the price of two cups of coffee!

Other recommended academic phrasebanks to keep on your desk for reference:

  1. The Only Academic Phrasebook You’ll Ever Need: 600 Examples of Academic Language

Academic writing books for specific disciplines

Many books are tailored to a specific research topic or field or discipline which is extremely convenient and important for you since you learn what is expected in your particular field of research not just in a general academic context or discourse.

NB: The following list covers very few fields and is just a taster of what this category looks like  – you will find tons of these for a lot of disciplines – please invest in one of these books that pertains to your particular field:

Computer Science

  1.  Writing for computer science by Justin Zobel – this is very essential book for the computer science folks who are writing essays, reports, dissertations or theses. It is a recommended reading in our school for all CS students at all levels: undergraduates, Masters and PhDs.

Nursing/health/Medical fields

  1. Academic Writing and Referencing for Your Nursing Degree by Jane Bottomley & Steven Pryjmachuk
  2. Critical Thinking and Writing for Nursing Students by Anne Harrington & Bob Price
  3. Writing Skills in Nursing and Healthcare: A Guide to Completing Successful Dissertations and Theses by Dena Bain Taylor


  1. The Principles of Writing in Psychology by T. Raymond Smyth  & Thomas Smyth – Holy cow! this book is just amazing! it is an encyclopedia that covers scientific writing, referencing, styles, academic standards, writing essays, literature reviews, reporting statistics. It covers also academic papers writing and publishing, qualitative research, how to present figures, tables etc… among tons of goodies – The book is dedicated for both undergraduate and postgraduate. I totally recommend this not only for the Psychology folks but for all disciplines.
  2. Critical Thinking in Psychology by Robert J. Sternberg, Henry L. Roediger & Diane F. Halpern
  3. Effective Writing in Psychology: Papers, Posters, and Presentations by Agatha M. Beins & Bernard C. Beins – an amazing book for writing papers. I know a supervisor who recommended this book for her PhD candidates to read.


  1. Successful Legal Writing by Edwina Higgins & Laura Tatham

Books that teaches you how to write papers and publish

  1. Writing Scientific Research Articles by Margaret Cargill & Patrick O’Connor

Citing the right way

Trust me! A lot of folks in academia do not know how to cite well. Many do not cite well out of being sloppy not for any nefarious intention. Please learn how to cite well. There are very good books that teach you how to do that. They teach you many citation styles  such as APA, Harvard, IEEE, MLA, Turabian, Chicago etc…. They teach you also how to cite per example an online video, a website, a lecture, an academic journal paper, a book, a chapter, a DVD, a phone call or an interview with an expert in your field etc… There are zillions of things to know. In some cases, researchers in a certain field need to use a certain recommended citation style so they need to consult a referencing style book covering only one particular citation style (For example: APA for the Psychology folks).

IMPORTANT NOTE: relying solely on referencing software such as Mendeley or Endnote or Zotero or on Google Scholar or the myriad publishers websites to export for you a BibTeX file or an RIS format file or whatever other format for a citation yields in many instances to incomplete and wrong citations (from my experience) – This is why these books are extremely important. 

  1. Cite Them Right: The Essential Referencing Guide by Richard Pears & Graham Shields
  2. The complete guide to referencing and avoiding plagiarism by Colin Neville

Finally on this topic, there are some websites that help you to cite correctly according to a specific citation style such as Cite This for Me and citefast. The problem with these websites is that the method used is too primitive or too manual since you have to fill out manually a lot of necessary fields for citing a resource compared to using a referencing software such as Endnote or using BibTeX/BibLaTeX export features in scholarly search engines and publishers’ websites but the advantage is that you know that you are following exactly the citation style that you need and that your citations are correct and complete.

J – Master NOT being isolated and the art & science of presenting research

You should aim to finish the PhD while keeping your sanity intact! Otherwise you will develop a lot of mental health problems, the least worrying of them would be “a severe depression”. If you ever found yourself completely isolated in a PhD in other words, you can not talk to anybody about your research or your problems or your frustrations, this would be bad, I mean very very very bad!!! This is a situation of real concern. If you think it is not: it is either you are dumb or whoever telling you not to be concerned is dumb.  Don’t live in ivory towers! Now of course you will ask why? In a lot of times during my PhD,  I wished I have talked to an academic in my department or a colleague or attended or submitted an academic paper before the time I did since other people i.e. experts and even normal academic staff can definitely point out things in your research that you and your adviser did not see. Things that might be better or maybe sometimes they could mention methods that are easier or more direct. Remove the idea that you adviser is always right or smart because this is not always true. I used to get sometimes the reality questions from people of the form: “why did you use X not Y? Y is a lot direct and easier?”. It is not like X was wrong but it was the method that takes more time than Y per example.

Attend conferences, lectures, seminars and events in your school or department. Create a network of friends from PhD candidates, postdocs, research staff, other supervisors and collaborators in your school and outside your university. This is something that only you can do. They can be of great help to you when you have concerns or problems with you primary supervisor or in your PhD. This group of academic staff, postdocs and candidates can comment on the content of your reports, on paper drafts, on chapters of your thesis etc.  They can check whether there is a logical flow of ideas, whether the overall argument is cogent, whether the structure makes sense etc. They can attend mock presentations for conferences, or help you in a mock Viva Voce etc.. They can give you tons of useful and needed feedback and advice.

In a PhD, outside befriending your supervisors, I advice you from my experience to always befriend a librarian (trust me!), an IT specialist, an academic language specialist, other supervisors & staff, PhD candidates, postdocs, at least a staff who deals with your university’s procedures and regulations and at least a UK lawyer or someone who is studying UK law 🙂 Of course these would remain valuable friends outside the PhD when you finish.

Ask a lot of questions and do not let your head being eaten

You need to be very inquisitive and ask a lot of questions to a lot of people either about their research or your own research. Of course no one will do your PhD for you, you will do it. On one hand, asking a lot of people about what they think of your own research and about the problems and challenges that you are trying to solve, will give you the opportunity to have a lot of advice and good feedback both technical and non-technical and a more appreciation of what you are trying to do. It also makes you see the bigger picture of things i.e. a bird’s eye view. On the other hand, when you ask people about their own research, trust me!, this is not a futile exercise, it teaches you the skills necessary to look critically and objectively at your own research and how to defend the decisions you take from seeing how others do that when asked. In addition, it makes people remember you per example in conferences especially when you ask very good questions. Also there is something important that I did not mention yet. Talking to people (staff and students) is essential when you have problems with your supervisor or your PhD, since many are happy to give you useful advice on how to rectify the matter from their own experiences before the problems become acute or before it is too late to do anything about them. In contrast, isolation leads you nowhere!

Some advisers become clueless especially when you become more knowledgeable than them in your particular area especially at advanced stages of the research so you need always experts or near experts to criticise your work which is very good for you. Clever PhD candidates normally eats your head figuratively speaking 🙂 with their bombardment of intelligent questions and criticism and this is a great and a healthy skill. You need to be like that. In other words, you need to be the initiator of intelligent questions not always the recipient of them by others.

When your supervisor or adviser ask you why not do X now? Do not follow on command like a faceless idiot. This might  be costly to you. Ask why you think I should do X now? is it really in line with my PhD “story” or “arguments” and how? Do I have the time & resources now to do X? What do you think? Has it been done before in the literature? if yes, why are you suggesting it? Shouldn’t I do an extensive literature review before? Trust me! a lot of advisers sometimes try to look smart in front of you by suggesting to you fancy or irrealistic ideas that either are in the literature hidden deep somewhere already or are un-achievable with the time frame and resources you have. You the poor guy or gal think well if the supervisor is saying it, that means it must be true or useful to my PhD. A PhD programme is designed to make you a real academic and a real academic is a person who asks always a lot of questions and is suspicious by nature. An academic is a skeptic by nature. The moment you stop being skeptic is the moment that you become an official idiot!

Presentation skills

You need to master giving a lot of talks for conferences, or in front of posters. A lot of universities in UK organise every year poster sessions for PhD students. You will present regularly your research in front of your supervisors and other fellow PhD students. You will present it at conferences, workshops, in PhD yearly reviews and of course at the end in your big day: the Viva Voce…. So being shy and afraid of public speaking might disadvantage you.  Learn the art of preparing an “elevator pitch”. This is very essential. You need to be confident. When you talk in front of any audience, you should talk with the confidence of an authority because after all, the research that you are conducting should make you an authority. Right? All you need is a healthy dose of confidence.

Before I move on to the next advice, according to the Ph.D Symposium ICSM 2009, it is advisable that you present or “explain you PhD topic to at least one person each day“. Now of course I think this might be a bit exaggerated but you can set goals like each week you need to present you PhD to at least one person.

I believe it is silly to say master the usage of a presentation software such as Apple Keynote or Microsoft PowerPoint but there are still a lot of people who do not use these software to their fullest potential. There is nowadays a new famous kid on the block which Prezi, a presentation software that allows you to express motion, zoom and spatial relationships among your ideas in a presentation.

As a general advice, avoid flashy animations or anything that can distract your audience in your slides. Keep them as simple as possible! You should use very short text in the form of lists, a lot of images, and a lot of tables. The point that I am driving home here is that the content on the PowerPoint, Keynote or Prezi presentation slides should be just a reminder for what the presenter should talk about. You do not want your audience spending time reading the slides instead of focusing on what you are saying. Avoid monotone voice when presenting because this will send your audience to their beds.

We are seeing a new trend these days in many workshops and some very few conferences that involves forcing attendees to present what is dubbed as “PechaKucha” presentations. A PechaKucha is creating 20 slides with only 20 seconds each i.e. 6 minutes and 40 seconds in total. The point here is to force the presenter to get to the point quickly and efficiently instead of rambling around on each slide.

It all about connections!

Conferences are great places for acquiring connections that could help you out in the future in different scenarios (collaboration, applying for future academic positions, choosing externals etc..). Events organised by your school is also another venue.

One cool thing about the life in UK universities, I guess also it applies to US universities, is the availability of students’ societies. Join the societies that you are interested in. In UK, all prestigious universities have tons of student societies.

Of course, don’t make this in detriment to your PhD or the necessary time that you should be allocating to your research. Student societies could be a good useful entertainment while giving you the opportunity to meet a lot of students and staff members (not only students). Students’ Societies spans from being social or hobbies oriented, to spiritual, to philosophical, to even magical, to religious, to political and finally to societies by invitation only (sort of what the paranoid masses like to call them secret societies LOL!), you will probably not find something similar to the Skull and Bones or Scroll and Key Yale University societies but there are some that require that either one or two members should invite you to join.  Fraternities and sororities are a lot more prominent in US universities than in UK universities.

K – Don’t miss out opportunities to teach

Help out in teaching, demonstrating, tutoring and marking modules taught by your school. Even help out as an exam invigilator! why not? They normally need a lot of PhD candidates to help out with semesters’ modules. You might have to take a few workshops and quick courses before you are allowed to do that. This will give you extra cash and income. Bare in mind that if you are an international student, you are only allowed to work a fixed number of hours per week. At the time of writing, a tier 4 general student visa allows you to work only 20 hours per week.

Luckily in UK, teaching is not obligatory which might be the case in other PhD programs around the world. I love teaching. Teaching a course obliges you to master it from my personal experience. “The best way to learn is to teach“. In a similar vein, do not overdo it because it will definitely affect your research and your PhD which should be your main objective. Always strike a balance. You need to discuss teaching with your supervisor. The supervisor is responsible vis-à-vis the university or the department to not let you work if your PhD progress in not good. You and your supervisor might need to sign a certain form.

Finally I can give another advice here. You can propose a short course inspired or based on your special area of research. A lot of universities in UK encourage such initiatives. This might give you a change to share your research with a wider and more divers audience and will give an extra income.

In addition, a teaching experience is very good for your CV for both academic and non-academic positions. Even if you do not want to continue in academia, it will show prospective employers that you have many transferable skills.

L – Master doing only ethical research

I did not add this section in the initial draft of this article because I thought this should be self-evident. I decided to include a bit of advice here.

Avoid all forms of plagiarism. Many institutions and publishers use massive software such as Turnitin or urkund to detect plagiarism. Publishers such as IEEE use also a tool called CrossRef Similarity Check. Plagiarism is a great taboo and a serious offence in academia. A taboo with “no return back” or no “second chance”. So it is that serious!

Furthermore, avoid any type of fabrications or falsifications of research results. This is another serious offence. In addition, your experimental methodologies should allow others to reproduce the experiments and get the same results. Well!  the main aim of writing methodologies is reproducibility. Do not make grandiose claims or conclusions.

Stay always up to date with the literature so you do not end up duplicating other people’s work, unless if the main objective of your research is to duplicate certain experiments. This is because the most important criterion for a PhD examiner to tick or check on the examination form is the criterion of  “originality” or “original contribution to knowledge“. You might also be replicating at the start of your PhD, other people’s work to get acquainted with the topic and that is absolutely fine and even recommended. If you discover something wrong with those experiments that no one did discover before: this is still considered an original contribution. Per instance, you could have discovered a faulty dataset or a faulty experiment.

I should mention also that Literature Reviews (LRs) are not just for first year PhD candidate. Although some PhDs in humanities and social sciences might require that the LR is done in the first year i.e. submitted before the actual work starts.

There are other stuff you might see in academia which are completely unethical but veiled under a pile of hypocrisy. On example is putting names of academics that did not contribute anything to a paper or they have contributed very little. This is called abuse of power [5]. Students are forced by pathetic supervisors to put names of other academics as a form of “scratch my back so I can scratch yours”  or “quid pro quo” (I discussed this phenomenon before).

Anyone who did not significantly contributed to the writing of your paper should be put in the acknowledgement section with no exception. Even a small contribution to writing the paper such as creating graphs or pictures should be put in the acknowledgement. This is not me talking; this is the IEEE publisher requirements. I am sure all other publishers have similar criteria.

IMPORTANT: Do not put names of people on your papers that you did NOT collaborate with, or names of people that did NOT contribute anything to your papers or your work even if your supervisor tell you to do that for political reasons.

A lovely Professor in my school recounted his experience of his first time examining a Ph.D.  This was from more than 20 years ago. He was the internal examiner i.e the examiner assigned from the institution itself. He looked at the candidate’s thesis and found a lot of published papers for each chapter with 5, 6, and 7 authors on them including the candidate’s name. He did not know what to do since this literally means the candidate did only a small part of the work or his/her role might be limited in the published work. So he asked the candidate in the Viva Voce concerning that and the candidate said frankly that he wrote the entirety of all the papers and did all the work and he just put those names for departmental political reasons.

The point that I am driving home here is that it is more difficult for you to defend your research in your Viva Voce using published papers with too many authors on them even if your name appears as the first author. Each author share the contribution: adding more names makes your slice smaller! 

More authors’ names on your papers literally means that you have contributed less than what you should have contributed for a PhD. Please always bear in mind that the two main requirements of a PhD are to have (1) substantial (2) contribution to the knowledge of the field (two requirements here: the scope requirement i.e substantial (lots of it) + original contribution to the literature of the field). There are other requirements of course, please refer to this article for more information.  There is nothing wrong with collaboration and this is required and very beneficial to you  but please bear in mind that having many names on your papers will eventually hinder the “defensibility” of your work at the end of the PhD.

So appeasing your supervisor would have a very detrimental effect on you later. You supervisor has no say in your final PhD assessment of the thesis nor in the oral defence whether you appease him/her or not, whether you are loyal to him/her or not, whether she/he likes you or not. I am talking about the UK system. Of course, you should aim to have the best professional relationship with your supervisor based on mutual respect but regardless of that relationship, he/she is inconsequential to whether you get a PhD or not and I think I made this point very clear throughout the article.

If your research involves assuring anonymity of participants in your experiments, make sure you anonymise correctly. Use pseudonames for institutions or organisations, and use codes or numbers instead of names for participants. Do not forget to blur the faces of everyone in your photographs before you include the photos in your papers or your thesis especially if you are allowed to take photos of your research setting, equipment, and the people involved etc…Same advice applies for videos. The only way you can keep faces un-blurred or names revealed only when you have received written consents from the people involved. Don’t ask for verbal consents. They are not binding & they are useless and they open a pandora box for law suits. Preserve confidentiality if required.

Avoid any conflict of interests as I discussed in detail previously. Conflict of interests could be of many types: commercial, personal etc… Any potential conflict of interest should be disclosed and put into rigid defined boundaries as the university’s policies mandate.

You might be asked at a certain stage to peer-review academic papers in your field. Always preserve the confidentiality of what you review and always declare any conflict of interest. Always follow the exact guidelines & standards of the journal or conference proceeding. Don’t piss off editors by being late or giving a biased or rushed review. When you are late, you delay the whole publication process. You have to be objective, fair, professional and you have to deal with the work that you are reviewing with complete confidentiality while preserving anonymity.

In certain competitive scientific fields, few cases were recorded of paper reviewers or referees stealing the ideas of the academic papers they were reviewing. This is an extremely unethical act and it is illegal. It is plagiarism
after all. It has sent few bad academics to courts. Delaying referees’ reports in order to prevent the consideration of certain academic papers for political or personal reasons is also both unethical and illegal. You should never review a paper written by an author that you have an issue or personal vendetta with.

Many types of research require ethical approval  from your university or your school ethical committee for all experiments or studies that involve human subjects and animals. You need to be aware of the whole process so you can plan ahead. These things can take a lot of time. Ethical approval for research projects that involves dealing with pupils or exterior governmental services such the NHS tend to take the longest time for approval. You do not want to spend time waiting for such process so keep that in mind.

Also I want to mention that simultaneous submission of a work to more than one journal or conference should be disclosed. Usually journals and conferences do not allow the same paper to be submitted at the same time to different places and your paper normally will be rejected prima facie. There are few exceptions.

M – Master the art and science of publishing research & take any opportunity to practice professional paper reviewing

Doing a literature review is an essential skill to have. See a well received article that I have written on how to do that systematically. Everything I did in my PhD was in the form of numerous full reports in LaTeX and MS Word. These were done in order for example to document an experiment conducted or to document a tool that was developed or to survey a topic. Each one of these reports contained a complete structure of an academic paper: Abstract, Introduction, Literature Review, Approach/Methodology, Results, Analysis, Conclusion. Always follow that structure! This traditional structure has a name and it is called the “logico-deductive structure“. Even if you are developing a software, try to follow similar structure with the reports you will write and send to your supervisor. This will make it easy to publish the report if you and your supervisor are happy with the quality and suitability of the report.

Always aim to publish any work you do in the Ph.D from year 1. Original contribution to the knowledge of the field is the most essential requirement for any PhD. Your Ph.D work has to be substantial meaning lots of work of importance and impact, written and well explained , of good academic quality worthy of a PhD and should be publishable i.e. (1) lucid, (2) scholarly,  (3) substantial and (4) contributes to the knowledge in the field. Please refer to an article that I  wrote on what are the criteria of a PhD, thesis assessment and Viva Voce preparations. Ideally Ph.D candidates aim to have at least one peer-reviewed publication for every chapter. Publications tell the examiners of your Ph.D that your work was peer-reviewed and was found worthy by other scholars in your field. This is why publishing is extremely essential to the success of your Ph.D. You can still get in UK a PhD without publishing but with publications your defence is a lot easier in your Viva Voce. In some educational systems around the world, you must publish in order to get a PhD.

If you have a work worthy of publishing, and you do not know the right venue, a lot of publishers have recommender systems such the IEEE Publication Recommender. Probable venues for you to publish are nothing but the relevant and most prestigious journals and conferences that you have cited in your reports or in your literature review. Also your supervisor and other academics in your department can suggest a lot of venues for you to publish in.

Always know what are the best journals and conferences in your field of research. I mentioned this in my literature review article. Make these journals or conference proceedings your digest reading throughout all the years of your PhD even till the end. Someone should be able to ask you on the spot: what are the top 3 conferences & journals in your specific domain of research? and you should have a quick answer to that. The aim is that you would hopefully publish in such venues.

Your adviser should help you in this regard. Don’t be shy to ask simply because s/he is supposedly more experienced than you and knows which conferences or journals are worthy and which are not.

Make these top journals or conferences your daily or weekly digest from the start of your PhD and continue this behaviour through all the years of your study. You should subscribe to these conference proceedings or journals, like you do with your favourite car magazine or gaming consoles magazine. Many famous journals and conferences have email alerts and social networks’ alerts that inform you when new issues are published.

Another good advice I can give you here is that you should get an ORCID account in order to protect your academic identity and your published papers and your future publications. An ORCID is a persistent unique identifier for researchers. This is because a lot of people have similar names. I do not want someone with the same name like mine to claim my papers. All publishers now ask for an ORCID to link your submitted paper to you and only to you. Another reason for an ORCID account is the fact that you might be now at University X and you have a profile there but later you would move to work in University Y. Having an ORCID, a Google Scholar profile,  Research Gate and Academia profiles is very essential to preserve and protect your academic identity and to showcase your publications.

Concerning academic paper reviewing, Fong [4] strongly recommends postgraduate students to find opportunities to practice professional paper reviewing. Supervisors who sit usually on conference program committees can help greatly with that. You might be contacted by a journal editor to review papers especially if you have published before. Editors are always looking for qualified reviewers. Usually editors approach PhD candidates that have made an impact in previous publications. As a PhD candidate, it is good to be proactive and to meet all the editors and program chairs when you attend conferences.

Learn how to write a referee report. A Referee report is a report which recommends for or against accepting a paper and it lists all the suggested changes that the author(s) of the paper should do. In order to review a paper you should know the criteria required for saying whether a paper is of scholarly quality and whether it has made sufficient contribution to the field. You also need to match the quality of the paper being reviewed with the quality of the journal and with its acceptance rate.

A paper usually contains what is dubbed as Least Publishable Unit (LPU) or called also the Smallest Publishable Unit (SPU), or Minimum Publishable Unit (MPU) which is the minimum novel & sufficient contribution to the field that is worth to be published in a single paper.

Please only review academic papers pertaining to your specific topic of research and review only legitimate articles in legitimate conferences and journals. A lot of academics get bombarded on a daily basis by spam reviewing emails requesting review from low quality or spurious conferences.

Peer reviewing academic papers constitutes a great addition to your academic CV and it helps you learn how to be critical and it helps you also by consequence to address criticism in your own PhD work. In a sense, this activity strengthen your critical mind. It is also a good preparation for the Viva Voce and PhD thesis assessment. A PhD thesis assessment is similar to a journal peer-review process but on a much larger and rigorous scale.

N – Master managing efficiently your PhD and your time

You should manage your supervisor as I mentioned previously and take advantage of his/her valuable time and expertise not the other way around. The supervisor/adviser is just a human resource dedicated to help you on your PhD journey so take full advantage of that and appreciate the time given to you.

There many useful ideas for managing the PhD itself. The following are few ideas that help you manage you work during a PhD:

Backup your data continuously

You never imagine how many candidates I know who where too sloppy in not backing up continuously their data, reports, thesis etc… I have one  PhD candidate in the school of chemistry asking me how to retrieve desperately a report which he stored on a USB memory stick that does not work anymore He does not have any other backups. We tried putting it into a bag of rice and other crazy stuff using both software & hardware techniques but nothing worked. Why be like that? You have plenty of free online backup and cloud-based storage services such as Dropbox, Google Drive, Microsoft OneDrive, iCloud among many others where you can store backups. Invest in buying external Hard disks, and USB memory sticks. They are very cheap these days. Continuously backup up your data in many places. There are services and software and operating systems tools that help you in doing that automatically and intelligently. Institutions normally have backup systems for staff and students so ask your University IT services.

You can also create as a backup for emails received, a mirroring or a redirect private email using Yahoo, gmail etc… in order to redirect all your university emails to it. I have discussed before the importance of emails in terms of documenting minutes of supervisory meetings, drafts of papers etc… Bare in mind that after your graduate,  some universities remove your email account,  while others might give few months to backup things and delete private data. The university of St Andrews (my uni) gives all alumni access to their email accounts for only 6 months from the exact date of their graduation.

Create and populate a research diary/notebook

Go grab a big diary or notebook or better download one of the digital tools online or even create a  digital document on your PC or on cloud services. This diary will play the role of storing all your research ideas, the wrong approaches, what worked and what did not and always try to write the “why” of things, write also the lessons learned.  It could contain decisions, ideas, expectations of outcomes, papers that you have read, sketches of algorithms, mind maps, code versions, theorems, sources of data, experiments , sketches of proofs etc. Trust me it will help you a lot when writing your thesis later. This diary will give you a lot of material to defend your arguments, methods and findings.  You should know that you have to defend in your Viva Voce every single decision you have made (check out the article I am writing on Viva preparation and what examiners really assess in a PhD thesis). Try to always record all research activities in a chronological order as much as possible.  It is funny when you reach the end of the PhD and look back at the beginning and laugh!! The diary should be reflective and should be critical of yourself and of performance or progress so do not flatter yourself. Do not worry! no body will see it if you do not want.

A full-time PhD candidate is expected to work 40 hours a week. It does not matter how? In many PhDs, it could be from Monday to Friday from 9:00 am till 5:00 pm or 10:00 am till 6:00 am. Consider a full-time PhD as a full-time job. The beauty of the PhD is that many of them are flexible when it comes to time and that is scary because you might end up working too much on one day (12 or 14 hours) and too little on another (ex: 4 hours). It is good to learn how to manage your time efficiently so you can have also a social life that will keep you sane.

Time is very precious!!!

Don’t let father time work against you. The old man with a scythe and hourglass! The teacher of bad experiences!

Plan ahead everything! Plan your milestones, your deliverables. Plan the PhD yearly reviews, the poster sessions, the deadlines for submitting papers to main conferences & journals etc…. Plan the preparation for the Viva, the date of the Viva, the corrections period after the viva. Plan also the time to look for a job whether in UK or abroad. If you are an international student, you are in an additional disadvantage since you have to fit perfectly all this planning in your student visa time. Not planning your time in a PhD is a sure recipe for failure.

All graduate industrial jobs for students and even for PhD candidates have very strict fixed deadlines every year.  In similar vein, lectureships and postdoctoral positions have very strict deadlines. For instance, big number of companies in the UK industry have strict deadlines in the Autumn of every year (September, October, November). They also have a fixed quota of people to recruit. Bare in mind, here I am talking about just the deadlines for the online/email job applications so even if you are lucky to be picked up, the recruitment process itself is in many cases especially in prestigious companies or firms, so long-winded, and so sickening that it could easily spans months (some even 6 months): doing psychometric tests, 1st round interviews, 2nd round interviews, days at assessment centers.

I know this article is dedicated for the pre-PhD or prospective candidate and for the first year PhD candidate but I wanted to mention post doctoral jobs  here i.e. in the skill of “managing efficiently your PhD” so that candidates keep this in mind especially when they reach the last year of their degree.

If you are or will be a PhD candidate on a student visa, you have to factor the visa time in your planning. You must plan your PhD very well both in time and in money: the PhD proposal period, the settling down period, the different PhD milestones i.e. yearly reviews and the submission to most important journals and conferences in your field. You also have to dedicate enough time to the choice of the examiners, to the thesis submission, to the Viva Voce preparation, to the Viva itself, to the period of thesis corrections  which usually takes from 3 weeks to 3 months if things are OK, otherwise maybe plan a period for an appeal. Furthermore, plan for a period to look for a Job i.e. applying online, tailoring your CV and your covering letters to the specific positions and attending interviews.

An important advice for international students: there is for all PhD graduates, a 1 year visa called the Doctorate Extension Scheme (DES) which allows you after you finish your PhD to search for a job in the UK and to work as many hours as you want without access to public funds.

Evidence all your PhD skills!

Create an academic CV-like website to showcase and list all your talks, presentations, all your publications, your teaching experience among others. It should contain a portfolio of all the projects from previous degrees, from academic positions and from the PhD. Whether you end up in an academic or an industrial role, all employers ask for “demonstrable” or “proven record” or “proven experience” of a certain skill or technology. You need from the start of the PhD to “actualise” everything you do or learn. Skills have to be demonstrable, evidenced and proven to any prospective employer. The mere fact of stating something on a CV is literally worthless these days unless employers can see tangible proof. As much as possible, reify your somewhat “theoretical” or “abstract”  or “hard to prove on prima facie” skills such as communications skills or leadership skills etc… As a prospective employer, I would say why the hell do you think you have “leadership skills” and laugh a bit about it? You might convince me that you have those because you have founded a student society, a club, a political party, or took a managerial position etc…  For all Computer Science/IT folks: I strongly advice you to document and present all your developed tools on GitHub and to create an online portfolio of all the applications developed and all the projects completed during the PhD. Create a gallery-like website containing abstracts explaining your projects and with links to source code, data sets or executables. You can then later give a prospective employer a URL to your online portfolio. Every single skill on your industrial or academic CV should be evidenced. If you say you know the R statistical language, an academic or industrial employer needs to see actual complex R scripts or tools written by you. If you say you have a teaching record of X years, you need to evidence that via relevant references and you need to present concrete achieved results such as good student feedback scores or comments or recommendations or nominations.

Whiteboarding the thesis

One beautiful soul not from the realm of mere mortals has taught me this technique and since that time I am indebted to her. She do this normally with all her students. The activity is called “Whiteboarding the PhD Thesis“. I advise you to do this technique with your supervisor either before you start officially writing you thesis or just before the submission. The idea is to use a big whiteboard and write your chapters titles in sections of the whiteboard and then follow your arguments under each chapter. The aim is to see how your argument flow during across your thesis chapters. There is an important requirement for the PhD that examiners check on their assessment forms which is the coherence of the argument.

This method is also used at the start of the PhD in order to formulate a plan for all the PhD. A plan both in time or in research objectives/questions! You need to formulate a PhD story, a flow, a consistent and coherent narrative…..

Recall from the section describing the requirements of a PhD, that we should have a ‘coherent story’, a ‘coherent and consistent narrative’, a ‘coherent corpus’ tackling a specific topic in depth not in breadth. This is called the coherence and consistency requirement. This also applies to a PhD by portfolio (meaning submitting a set of published peer-reviewed academic papers that follow a coherent narrative which is accompanied by a written introduction and conclusion to pull everything together).

Another useful technique that is more relevant to the final stage of the PhD. The technique is called the backward reasoning technique taken from Saunders et al. [3]. This technique can be used in a thesis whiteboarding session. You start by stating clearly all the conclusions and contributions of your whole PhD research then from these you would say how they are based on your findings (from experiments, tools developed etc…), then you would be probed on what these findings are based upon. This will lead backward to the method(s) used (what?how? and why?) which should be based on a research strategy that leads to valid and reliable findings. The research strategy should be built upon a critical literature review and upon the research questions and objectives. As you can see in this technique you move backward from the final research outcomes of your whole PhD to the beginnings and you check if everything is cogent.

O – Learn how to draw professional diagrams, infographics and other research media

I know what you are thinking! You are probably thinking is this really a skill for a PhD? The answer is YES. Well it is just for a PhD but for any research degree. You should learn how to use drawing software such as Microsoft Visio, Dia Diagram Editor, Edraw Max Pro or any other tool out there. Some people use Adobe Photoshop or similar fancy image editing tools and that is fine! Others  use Microsoft Word, Microsoft PowerPoint, Google Docs drawing tools, Inkscape, GIMP or even Paint. I found from personal experience that Edraw Max Pro is an amazing paid tool that can create many types of useful diagrams. I also use which is a good free online alternative.

Whatever tool you choose to learn, make sure you are able to draw beautiful and professional publication quality diagrams. You might need to draw design diagrams of a software if you are a computer scientist, or you might want to visualise an idea. You should aim to impress your examiners/reviewers with beautiful, meaningful and professional looking diagrams in your PhD thesis!!!  Any tool will do the job as long as you master it and know how to draw what you need. It is better to learn that before the PhD since you do not have time to spend on learning the tools during the PhD.

You need to know how to draw Gantt charts and Gantt tables because you will use them definitely in PhD yearly progress reports that you will submit at the  beginning or at the end of each year. Some universities only require review reports to be submitted once at the end of each year of your PhD. You will need Gantt charts also to communicate timings of research tasks and milestones between you and your supervisor. A famous project management software that allows you to draw Gantt charts is Microsoft Project.

In a similar vein, there are other types of diagrams that are very important and very useful in the PhD such as concept maps or mind maps and relevance trees . They help you brainstorm ideas with your supervisor even I recommend including these diagrams in official reports submitted for yearly reviews.

In a nutshell, you need to know how to draw the following diagrams – trust me YOU WILL NEED THEM at different stages of your research and in the thesis writing period – please check that you know how to draw beautiful diagrams such as: Venn Diagrams (good to see them in the Introduction and literature review chapters, they situate the precise topic you are working on among parent topics per example), Organisational charts, Gantt charts and Gantt tables (for time management), Work Breakdown Structure Diagrams, Flowcharts, Concept Maps, Mind Maps, Relevance Trees, Block diagrams, Word clouds, Network Diagrams etc. and of course if you are doing a PhD in Computer Science, you might need to draw UML 2.0  or ERD  or SysML diagrams. As a CS researcher, you might need to know how to draw them and what usage they serve in the design of a software. You might also need in Computer Science to know how to draw software design diagrams such as wireframes.

It is extremely amazing if you know also how to draw professionally looking infographics. Please don’t use the cartoonised ones! You supervisor and examiners will love you! Infographics are used a lot in textbooks and if you can show you can replicate some or create something totally new. Bob is really your uncle I guess!!!!

I like personally to use It is free and on the web!!! It is amazing! I use it for all sorts of diagrams that I need.  All the above mentioned diagrams are available in web tool and there are many libraries for icons, symbols etc…  You will like it!

Word clouds (a.k.a tag clouds) are also used in qualitative data analysis and many CAQDAS tools can generate them for you. Please refer back to the section that talks about CAQDAS software and qualitative data analysis. Word clouds or tag clouds can also be generated by a lot of tools online. Some tools can generate them in the form of faces or shapes if you want.

You will need also to know how to draw professional looking diagrams for posters. There are definitely many poster sessions throughout any PhD whether in the department or school to showcase your research in yearly poster sessions or whether to showcase your research in conferences and workshops. PhD candidates use usually Inkscape, Photoshop or Microsoft PowerPoint to create posters. MS PowerPoint is the easiest software that you can use for such a task. Posters are usually of sizes A0 or A1 in landscape mode.

Also I want to add here in  few things. In some cases during your PhD you might need to learn how to create a movie explaining your research. A lot of folks like to use Adobe Premiere. You can of course learn how to use any video editing or video production software that you find suitable. You might also need to learn how to create a professional looking graphical abstract or even useful animations. These media would be ideal for presentation in conferences or even to be shown on a website that you create for yourself acting as an academic CV and a portfolio for all your research projects. While on the subject, I strongly recommend PhD candidates to create a professional web presence. It is very essential in my opinion. In addition it is good to have a blog where you document your PhD and write about interesting things.

Aim always to produce high quality diagrams for different scenarios (academic papers, thesis, PowerPoint slides, web sites,  big posters…) so per example: 300 dpi to 600 dpi are the recommended resolutions for anything that will be printed. You can easily decrease a resolution of a picture you have produced but it is difficult (not impossible) to increase the resolution of a terrible image or graph.

A Side Note – Important advice when using – You can export a PNG, JPEG, SVG, PDF etc. version of your diagram with the tool. I find from experience, that if you use the PDF version of the diagram, the one exported from, the diagram has a better resolution than the PNG or JPEG  especially when you include it into your LaTeX document whether for a paper or for the thesis or even for a poster. When you finish drawing what you want to do, always select everything in the diagram by pressing Ctrl + A and then go and do the export. Do not forget to select the “crop feature” when you do that. This will generate an exact image with the same size as the drawing not a whole white page with a small diagram in it.

Specialised Survey and Experiment Software & Services

In this section, I will present few famous survey tools that helps you create questionnaires. In addition, I will also present few interesting applications that PhD candidates use for conducting experiments in certain fields. This section is a work in progress. If you have any useful suggestion of any application that can be helpful to PhD candidates, please post it in the comments.

Specialised survey tools and services

Familiarise yourself with questionnaires’ online services or software such as SurveyMonkeyQualtrics, Snap Surveys, Surveygizmo, crowdsignal and Google forms among others especially if your research involves designing questionnaires and capturing responses from them.

Many of these services contain GUI components such as check boxes, sliders, text boxes, matrices etc… that allow you to design a questionnaire in an easy and fast way. There are usually templates ready to use. These tools also code the responses for you so that you can use statistical tools to analyse your captured data.

Specialised experiments’ tools and platforms

You might need to familiarize yourself with open source platforms that help researchers design experiments, collect data and analyse results.

  • The Touchstone experimental design platform
  • For Psychology, Neuroscience, economics PhD candidates there is a program called OpenSesame which allows you to create scientific experiments. You need the Python programming language interpreter to be installed on your machine. Knowledge of Python programming language is advantageous. If you do not know Python, please check the Computer Science Skills section for more information.

General advice

In this section, I have few advice that do not fit into a skill per se but are very useful:

  • If you are in the UK, please do not forget to take advantage of the SCONUL scheme that allows you to access the majority of universities’ libraries and borrow books from them. Your Eduroam (the world wide roaming for education & research) account will work automatically and you can access all the journals the same way you do in your own university library or in your university’s internal network.  It is ideal for the folks who live in one city or town that has a university and study in another university further away. A lot of PhD candidates are not even aware of this. In UK, there are many good public libraries with big collections of academic books. Check out public libraries near you and apply for a membership card that allow you to borrow books. Many libraries give you the card without any headache while others, especially if you are an international PhD candidate in Scotland, require from you a card called a National Entitlement Card. Don’t worry! it is very easy to get it from your city or town council. You need a proof of residency that shows the address and an ID such as a passport and that is it! Many of these public libraries have very good quite places to study. Trust me this is needed especially when main libraries are busy especially in exams periods and of course when you have an annoying talkative office mate! When I was doing my PhD, I spent a lot of times in numerous libraries across 3 universities and across two cities’ public libraries. I found that I can avoid writer’s block when I simply change the place of where I write and work.
  • Always always always run at least one pilot study or pilot test or a pilot experiment before you do the real study or experiment – It does not matter what you are doing, if you are a social science guy/gal or a natural science guy/gal, or a CS guy/gal doing user-based studies or you are measuring metrics programatically or doing an experiment in chemistry or biology, it is the same concept, do a small pilot experiment/study or a test experiment and then go through ALL phases: from the analysis (statistics…) till the reporting phase.  Take the opinions of your supervisors and other academics on the pilot study/experiment results, analysis and on the report. Afterwards, start doing the real experiment or study. You can not imagine how many PhD candidates that I know who rushed to do their actual experiments after they supposedly “designed them well” and then they realised in a very sad way that the experiments contained many errors in the methodological approach, or in the sample and so on and so forth… Pilot studies/experiments can identify potential biases and methodological problems. NB: It is recommended that the participants in pilot studies/experiments should be from the target population.
  •  Do not forget that you need to prove the internal consistency or reliability of any questionnaire so learn how to calculate the Cronbach’s alpha coefficiant. You should understand very well what the number is really telling you and correct your questionnaire accordingly. To learn how to do that please check this YouTube video using Excel and  YouTube video 1 & video 2 using SPSS. It is an easy process. There is also another method that is more complicated and sometimes it is more preferable which is the Exploratory factor analysis.
  • For sampling (ex: surveys/questionnaires/human experiments etc…) you might need randomization. There are many digital randomization services such as that can help you out. Statistical packages and languages such as SPSS, SAS, R and others have also many randomization functions. You can also go old school for whatever weird reason and use random digit tables but that is extremely archaic and primitive.
  • If you are doing users-based/participants-based studies, always have a good rewarding system (good incentive) so that people can spend the required time on your project without feeling they are participating perfunctorily.  The least thing you want is for people to give you fake or rushed data. Think of a good, attractive and reasonable rewarding system and do not overlook this. Chocolates and sweets are not really very attractive especially if you want a big divers sample. Friends would be happy to help but sometimes you need a bigger and more divers & random sample and a convenience sample is not an option. I found Amazon vouchers to be a good rewarding system, the vouchers do not have to be 50 pounds or something like that. I mean if you have generous funding, go for it!  Five or ten pounds Amazon voucher cards might be very useful for participants as a thank you and could attract a lot more for advertised studies.  Every funded project set aside money for a rewarding system for participants so you will not buy anything from your own pockets! A lot of people purchase stuff from amazon so getting a voucher discount of 10 pounds obtained from participating in a study that takes 30 minutes is not a bad thing for a participant! You can buy vouchers from Tesco mainly the big stores in UK or online from Amazon!
  • You might need to use a software that copy files across computers or from a client machine to a server or vice versa (FTP/SFTP). FileZilla , Cyberduck, and WinSCP might be useful tools for you to be familiar with.  There are others of course. You might need to learn to SSH to a server from a terminal. I am saying this here, since I have a friend who is doing a PhD in Physics who asked me to help him out SSH-ing to a server where he is storing his datasets. So this is not a CS exclusive thing! If you are on Microsoft Windows machine, you can use a popular tool called Putty. The usage of Linux and their terminals is not only for Computing Science folks. I know many people doing PhDs in scientific fields such as Physics, Astronomy and Chemistry who have many scientific applications developed only on Linux.  If you are NOT doing a Ph.D in CS but you still need heavy usage of Linux OSs for running  scientific or visualisations applications, please have a look at Skill 1 in the Computer Science Skills section for some useful suggestions of online resources, books and online video courses on the topic.

Questions that might be asked at one point or another in a PhD

1- Can I copy verbatim my published papers or how much verbatim can I copy from my published papers during the PhD into the thesis in the write up?

This is an extremely common question and I asked an esteemed source on copyright matters and got some good informative answers.

First thing that comes to the mind of a careful PhD candidate: am I committing self plagiarism? The answer is clearly no, since these published papers are part of the work done for your PhD degree, they contain your own ideas, experiments etc…  and the thesis is the culminating submission of your PhD degree. If you are using material from papers published outside the PhD maybe in another degree, that would be considered self-plagiarism unless if you are just citing or quoting material properly, that of course would be fine.

But hold on, there are two issues if ignored they will get you into academic misconduct (very serious trouble!).

First, acknowledging contributions of other authors if there is any: normally aside from the supervisor unless if the supervisor actually contributed to the work on any level. Mere supervision task is acknowledged in the thesis differently. Usually other members of the research team might have practically contributed to the paper so you have to exclude their contributions from the thesis and specify clearly in your thesis what have you contributed to the work published and what have you not. This is very serious! if you don’t do that you are committing plagiarism by attributing others ‘ideas and contributions as yours. I know also for a fact that in many cases, supervisors actually force students to put names of people in their department or elsewhere who have not contributed anything. This is an unethical behaviour that is very common. In this case no need to acknowledge anyone who did not contribute but always have concrete proof (emails, notes, audio recordings etc…) in case something comes up in the future. Have with you always the emails documenting all previous drafts, and the absolute initial report you wrote yourself and have sent to the supervisor(s). You never know! it is a hypocritical environment!  PhD candidates have to acknowledge also at the beginning of the thesis whether they used equipment or software of other research team members and in what extent.

Second, now acknowledging any contributions is out of the way, a lot of time even you might be the only author of the paper or the paper would have only two names: you and the supervisor with no contribution of supervisor other than the supervisory work: i.e. feedback & comments. The questions asked now are: how much can you copy verbatim or can you copy verbatim in the first place?

The problem now is a copyright problem with the publishers. I will hand the answer to very trusted source about copyright matters from a well respected Scottish university (the given answer is verbatim) –  The source wanted to stay anonymous and I  respect that but he was happy to give his consent on sharing his response. I should mention that I started the conversation with this esteemed source on copyright matters with the fact that it is advisable for the PhD candidate to review the copyright documentation he or she have signed with the publishers.

As you say, the starting point would be for the student to check their agreement with the publisher to see what they and their co-authors agreed regarding copyright and reuse. The agreement may allow them to include the published texts verbatim in the thesis without permission from the publisher or use of a copyright exception.

If however the authors assigned copyright to the publisher, and there is no clause in the agreement allowing reuse in a thesis, the student will need the publisher’s permission to include extended verbatim extracts in their thesis. I’ve included a template for a permissions letter that could be used as a basis for this. It’s our experience that this kind of permission is normally granted for the print thesis.

It’s also worth checking the publisher’s open access/self-archiving/author reuse (or whatever they call it) policy, normally found on their website, which sometimes state that authors are free to republish in their own future works, occasionally with limitations like ‘as long as it includes x% new material’ or ‘doesn’t constitute more than X% of the total new work’.

Agreed that it’s worth exploring using versions of the published articles held in open access repositories, but bear in mind that these may also have restrictions attached, so again you would have to check the publishing agreement and reuse policies before copying from these.

The student could copy shorter extracts without permission by making use of the “Illustration for instruction” exception (Copyright Designs and Patents Act, section 32, June 2014) – however this would be subject to the limits of ‘fair dealing’ which would not allow the extensive copying you suggest. There is no precise definition of what ‘fair dealing’ is, so I can’t give an exact figure on how much you could copy under this exception – roughly speaking I’d suggest that copying a paragraph might be considered ‘fair’.

Remember also that the thesis may eventually be published online in the form of an e-thesis – the “illustration for instruction” exception would not apply in this case as placing the thesis online is considered a form of publishing. Therefore explicit permission to copy from the published articles in the e-thesis would need to be sought from the publisher if the publishing agreement and their reuse policies do not allow this. It may be possible to embargo the e-thesis for a period of time, or edit it to remove the infringing sections, if permission cannot be obtained.

It might also be an idea to check with your supervisor and School or possibly the Pro Dean about attitudes to inclusion of pre-published articles in theses, and what percentage of the thesis can consist of such articles.

If the theses ends up being submitted through plagiarism software (e.g. Turnitin) it will probably be flagged if it includes substantial extracts from published articles – just something to be aware of.

Finally, re-writing the content of the published articles to include in the thesis would not be a problem from a copyright perspective, as copyright law applies only to the text and not to the ideas expressed within it.

As the esteemed source suggested: always refer back to your university policies, and ask the supervisor (s) or the Pro Dean who they will give you further advice. The safest of the safest approaches is to re-write the content of the published articles that pertain to your own contributions in case of papers of many authors and then include them in the chapters of your thesis as this is completely fine since you own the ideas and the work. Rewriting might be a daunting task. I know PhD candidates with 14 published papers. It makes sense to look at what is permitted by the publishing agreement or by your school policies before committing to rewriting a massive number of papers.

What you MUST know if you are doing a Research based degree or a PhD in Computer Science

You must know ALL the above major points or major skills because the aforementioned general skills are necessary for every discipline under the Sun and for every PhD under the Sun including a PhD in Computer Science.

It is pertinent also that you read at least one book covering research methods in your computer science field. Some sub-fields of Computer Science have canonized dedicated books for research methods to read while others unfortunately have only published peer reviewed papers. For example: concerning HCI/QoE research, books such as Research Methods in Human-Computer Interaction by Jonathon Lazar et al. or/and Research methods for human-computer interaction by Paul Cairns and Anna L. Cox and Doing Better Statistics in Human-Computer Interaction by Paul Cairns are all very essential readings for all the folks who will be doing any research in HCI or QoE.

It is essential that you refer back to sub-section “Computer Science methodologies” in the section “Skill G – Master Research Methods and Methodologies”. In this sub-section, you will find other essential readings for sub-fields of Computer Science at least what I know about. 

Unfortunately there is till now no canonized volume(s) for CS research methods summarising all the research methods employed in Computer Science as a whole similar to what we have in other disciplines.  It is not an impossible task but I would assume it is a very tedious one. It is something that Computer Science needs. We need some good Samaritans to summarise published peer-reviewed research found in the literature covering research methods espoused in numerous sub-fields of Computer Science such as AI, QoS, QoE, programming languages, software architecture, software engineering, data science, Health informatics, HCI among many others… We need good Samaritans that can compile them into one encyclopedia written in the aim of making it easy to consult by research students. In other words, it should be made easy to read by new researchers.

The following are skills specific for Computer Science (CS). It is pertinent here to mention that what will be discussed later are ONLY GENERAL skills for CS so you might need also to learn specific technical skills pertaining to your PARTICULAR research topic.

You might need per instance to learn Ruby and Ruby on Rails or I don’t know maybe learn .NET framework languages or Google Go or Docker or deep and machine Learning libraries such as TensorFlow, or scikit-learn or PyTorch. You might write scripts in tshark or Perl. You might be doing stuff in Apache Hadoop, blockchain and Ethereum or whatever is fancy these days. You might need to learn a functional language such as Scala or F#. I mean here it is extremely crazy, stupid and unproductive to have a fixed template of CS skills/languages to throw at you for the PhD purpose. I won’t do it and you probably won’t accept such a template.

That has been said, from my experience, the following skills are too essential and too general in the same time. They work on many dimensions if I can say that. Per example, I have put Python as a ‘ring to rule them all’ kind of language that you need to master and I believe this is becoming very true these days. Python is used to do statistics and data visualisation like what has been elucidated in skill ‘B – Master 1 or more Statistical Packages or Statistical Languages‘ so if you are doing a PhD you definitely need to visualise your data and do statistical tests so if you decide to embark on a Python journey, you do NOT need to learn R or SPSS or Stata for your statistics. This is a great advantage!!!. Imagine you are also doing a PhD in security,  or digital forensics or a PhD that involves machine and deep learning or a PhD that involves quant and financial data! bob’s your uncle, because there is a big probability that you need Python as a language for your specific research topic.

I am also assuming that you should have a very strong foundation and proficiency in at least few programming languages including a strong proficiency in data structures, algorithms, time and space complexity and design patterns. This is why I did not create a separate section for that. You should have experience building complex applications that are performant, multi-threaded or multi-process and scalable. This is of course something that is recommended to have for lower academic degrees and for all industrial software jobs not just for a PhD. Breadth of coverage is not ideal when it comes to programming languages, depth is what is really needed & recommended especially at this stage: PhD level. I do not see how learning 30 programming languages in a shallow way will help you in a PhD even if you are investigating programming languages paradigms.  Master one or two or maybe 3 programming languages to the core!  Usually, newbies state on their CVs that they have familiarity with many languages, while expert programmers love and stick with only very few languages in which they have obtained high proficiency by working on a lot of very complex or “meaty” projects involving advanced levels of efficiency, concurrency, profiling and performance optimisation and using many frameworks written in that language. Programming is all about practice, experience & exposure, the more you do it, the better you become at it. No amount of books or courses can replace practice. It is like going to the gym, you do not read books on the anatomy of your biceps muscles and on weight lifting  and then expect to develop some muscles from just reading about them. Being very good in programming comes from a vast experience  working on complex projects, from contributing to open source applications, from internships, from participating in hackathons, from working in industry for couple of years etc.  This is why I do not judge my readers, I do not know who is reading this article but I intended this article in the first place to be dedicated for the disadvantaged student from the disadvantaged country not for the spoiled folks with easy access to prestigious programming internships & to jobs in Google, Amazon, Microsoft, Apple, Palantir or Facebook.  I should also mention that programming is not only about programming languages or about your depth of knowledge in them, it is also more importantly about your problem solving skills and your ability to have an algorithmic mind. While on this subject, it makes sense and it is wiser actually,  and of course if it is possible, to choose a research topic for a PhD where you can take advantage of your programming expertise in certain specific languages. That has been said the PhD is also about learning new technologies and languages but the only real fear here is to get dragged into learning new languages and not to focus on your main research topic which lead to problems in your PhD progress.

It is also very essential to have a strong mathematical foundation as a pre-requisite for some PhDs. A wide range of research topics actually require that. Please ask your prospective supervisor or other PhD candidates or postdocs in your prospective research group about what topics you need to know. In other words, you might need maybe to take a refresher course or to do some essential reading on certain essential mathematical topics before you start your PhD. If you will be doing a PhD in programming paradigms, knowledge of lambda calculus, algebra and category theory are very beneficial. Example: functional programming languages are based on lambda calculus, algebra and category theory. Graph theory, Discrete Mathematics, Calculus, algebra (particularly linear algebra), probability and statistics are crucially important topics for prospective PhD candidates especially those that will be doing Computer Graphics, Machine learning and AI. Actually outside a PhD context, Math topics such as Discrete Mathematics, Lambda Calculus and algebra (particularly linear algebra) constitute essential knowledge for every computer scientist whether junior or senior. Not all PhDs in Computer Science require heavy math. Please check this before you start. There are some good online courses you can take on Coursera and Udemy covering mathematics needed for machine learning/deep learning.

Believe it or not, a lot of computer scientists even the most expert in the field always have what is dubbed in Psychology as the “imposter syndrome”. This is very normal since in Computer Science there are always new hip libraries, new fancy languages and new fancy terms that people throw around. I mean JavaScript used to be a joke simple language in the nineties now it is the language of the industry with new fancy frameworks. Some technologies look more intimidating than they are in reality. The truth is usually the opposite. This because of all the fuss about them in the media.

I strongly recommend you to actualise all the skills that you already have or that you will obtain during the PhD. So if you say you know the Python TensorFlow library, you must be able to demonstrate this tangibly for any prospective PhD supervisor or prospective employer via for example complex projects that you  have worked on in the past or a relevant industrial experience. I advice you to document and present all your developed tools during the PhD on GitHub and to create an online portfolio of all the applications developed and all the projects completed during the PhD. Create a gallery-like website containing abstracts explaining your projects and with links to source code, data sets,  or executables and to related published papers. You can then, when you finish the PhD, give a prospective employer a URL to your online portfolio. Every single skill whether on your industrial CV or on your academic CV should be evidenced. The skills that you claim to have and which look “abstract”, “theoretical” or “hard to prove on prima facie”, should be reified into something tangible and concrete that can be evidenced.

I have one last useful and very important advice before I elucidate the CS general & essential skills in the next sections. Please keep your saw sharp by staying current with up to date developments and best practices within the US/UK software industry especially when it comes to software engineering, development, testing and automation. Now, we are seeing a lot of cases where even the simplest jobs of software developer or software engineer require the inclusion of heavy DevOps in any job description. You are usually required to have knowledge in Continuous Integration/Continuous Delivery (CI/CD) tools such as Jenkins, ANT, Maven, Artifactory, Gradle, Sonar etc.. Furthermore,  jobs these days even for PhDs require knowledge of automation & configuration management tools such as Puppet, Chef, and Ansible. And as if things are not complicated enough already, you have to add to the stack of skills practical cloud computing knowledge on platforms such as Amazon AWS, Microsoft Azure or Google cloud. Big data technologies, Containers and Container Orchestration technologies such as Docker and Kubernetes also have their own share of your suffering. In addition, knowledge of agile & lean methods (SCRUM, Kanban, XP) is extremely essential these days. And people wonder why there is a lot of depression among Computer Scientists & IT professionals!?

My advice to you here is to always keep an eye from time to time on what is going on in the UK & US software industry in your PhD especially for senior jobs in companies such as Amazon, Palantir, Google, Facebook, Microsoft, Twitter, JP Morgan, Morgan Stanley, Goldman Sachs, Sky, Accenture etc… This is of course very relevant for the folks that are planning on doing a PhD and then on returning back to or starting a fresh career in the industry. You do not want to spend 3 to 5 of full-time PhD years living on Mars and then to return back to the software industry finding yourself very disadvantaged compared to other jobs candidates instead of having the advantage that you deserve. 

Keeping on eye on the software industry ground you and your expectations in reality, i.e. in the technologies that are being actually used, since many PhD topics are completely unrelated to what is going on in industry or at best they are very abstract and theoretical. In many instances (not all of course), academic research lives on a completely different planet. For IT/Computer Science industrial PhD graduates jobs,  many PhD graduates work in IT consultancy, data science, and machine/deep learning. In addition, game design/development, or Augmented Reality companies are looking for PhDs specialising in AI, computer graphics, virtual worlds, virtual reality, and augmented reality. Outside academia, PhD graduates in HCI or QoE tend to work as UX/UI/IA/IxD senior designers or researchers in companies.

As I said, not everybody do a PhD to stay in academia. I know that this article is meant for the folks that are either considering a PhD or they are already in the first year of their PhDs but I want to say that if your compass will be pointed toward graduate PhD industrial software jobs, please start searching for jobs at least a year before you finish your PhD. Pay attention! all graduate positions even for PhD graduates have fixed applications’ deadlines usually in the Autumn of every year (September, October, November, few in December). The process of recruitment itself takes extra months so plan ahead.

Please remember that it takes a lot of time to apply for jobs i.e tailoring your CV to the position in question and writing many tailored cover letters. Preparing for coding interviews or PM interviews takes also enormous amount of time. Without any exaggeration, it will take you several complete months to prepare well, and here I am assuming that you are working many hours daily. Here I am talking about the time taken to only refresh your memory after 3 to 5 years of PhD of common data structures, algorithms, time/space complexities, CS concepts and design patters: concepts you have learned in your Bachelor degree.  There are many websites that host many interview questions asked by major tech companies. Solving these questions takes a good amount of time. I strongly recommend you to bookmark these two GitHub repositories for later: John Washam – coding-interview-university  and Kevin Naughton repository.  3 PhD graduates from the most prestigious and oldest UK universities whom I personally know spent 4, 5 and 7 months just preparing for these types of interviews. They all applied to big 5 tech companies. I only hope you can remember this advice when you reach your last year!

Update 2019: Recently many UK universities are organizing workshops at the start of a PhD or of a postgraduate degree. Two types of workshops: the first is dubbed as “essential research computing skills”  a.k.a “software carpentry for research” and others called “research and transferable skills”. I believe this is amazing. So please take full advantage of those when you start. When I started my PhD we were not spoiled at all like the new folks now. So as an example from my school (Computer Science – University of St Andrews) they are now giving tutorials on Linux and shell scripting, Git and version control and Python). This article was written way before that but it seem they realized the importance of what is included in this article. Not to mention,  the school of CS  accept PhD students from all around the world with different academic backgrounds. Of course what is taught is these carpentry workshops is  not all what you want or even all what you might need but it is a very good start. Please check out also the websites:  The Carpentries & Software Carpentry.

Skill 1 – Master Linux/Unix, Usage of Terminals and Shell Scripting

Every reputable PhD in every reputable Computer Science School in UK, DOES require you to have a mastery of Linux/Unix and to have a mastery of shell scripting (bash, csh or tcsh, zsh, ksh…) or generally speaking to have a mastery of using unix-based  or unix-like operating systems.  Actually the mastery in this skill is required at the undergraduate level in UK not only for a PhD per se.

In the UK, Linux is everywhere meaning in all undergraduate CS modules from the start without exception. This is unfortunately and sadly not true for a lot of undergraduate and master degrees around the world especially in poor 3rd world countries or developing countries. In many countries, students are given a mere glimpse or taster of a Linux OS or any Unix-like OS in per example an Operating Systems course and that is it. The usage of such an OS system is only limited to such a course. Sadly nothing more than that. Unfortunately this is extremely bad for students coming from very disadvantaged countries to study in western European countries (mainly in the UK) or in the US.

In my school – the school of Computer Science, University of St Andrews – all our lab PCs are equipped with a dual boot between Microsoft Windows 10 and the Fedora Linux distribution.

If you have a Mac OS, you probably know that you can write shell scripts and even almost turn your BSD based Mac OS X terminal into a GNU Linux terminal through installing many additional packages from package installers such as  HomeBrew, MacPort and Fink.

Please bare in mind that I am using in this section of the article, the GNU Linux as a representative subcategory only.

By shell scripting, I don’t mean writing silly scripts. You should master scripting with powerful GNU tools like sed, awk/gawk, grep, tcpdump things of that sort, in addition to the traditional tasks like running applications from the terminal, compiling source codes, changing firewall rules, running and understanding network commands etc..

Learning how to script is very essential to automate experiments and to make them replicable by you or by other researchers in the future. Do not forget to document extensively your scripts so that you and other researchers can understand them in the future. It is also a good idea to share them on GitHub so that other researchers can use them in their own experiments.

I reached a stage in my Linux knowledge where I even listen to music from the terminal! Plus for some weird twisted reason, I reached a stage where I actually like or enjoy writing files in terminal editors such as emacs or vi or vim. In any circumstance outside a PhD, I would seek immediately a psychological treatment 🙂

As an advice for PhD preparation purposes especially for learners:  try to be on a regime of using Linux or whatever variation of Unix-like operating system for ALMOST ALL your tasks without exception in headless mode ONLY meaning just using a terminal with a keyboard and a screen with no GUI, no mouse or touchpad is allowed! for months (!!). Can you do that? Of course you would need GUI for some tasks but trust me this will teaches you tons of things that you can do: since you will be very frustrated to do certain tasks which forces you to learn how to do them from the terminal. You would be amazed of how much you will learn by this torturing technique!

Some academics even require their PhD candidates to move to totally higher level in Linux knowledge such as the folks working on implementing new network protocols at different level of the OSI layers stack or the ones who are working on the Kernel of Linux aiming to include a new feature. Usually here you should be able at a minimum level to configure and build a kernel in Linux and then to create your own Linux distribution from source code. Now some academics make it sound more difficult then it actually is.

I advice you if you are interested to checkout the Linux from Scratch & the Beyond Linux From Scratch projects that teach you how to build up a Linux, layer by layer from source code. I am thinking of creating a series of articles and videos or even a Udemy course when I have free time, on how to build and compile a Linux distribution from scratch because I always believe the best way to learn is to teach.

In a nutshell, long story short! basic average Computer Science Joe needs to know Linux and she/he must love it. This does not mean that you should not be OS agnostic. You should ideally love other operating systems such as Microsoft Windows and Mac OS or even mobile operating systems. I find all the folks who have a sort of a “partisanship relationship” when it comes to using an OS, immature and stupid!

If you happen that you need to work on Microsoft Windows or if your PhD needs this, I advises you to invest in learning PowerShell since it is a extremely powerful scripting language that do amazing things just like shell scripting in bash, csh & tcsh, zsh etc.. on Linux.

Specific Linux tools that I recommend every candidate to learn:

1) Learn to use a terminal multiplexer and how to script it. The most famous  are the GNU Screen command and tmux. Please checkout this article for more information. What is all the fuss about these tools? Why are they extremely important? They are super important because sometimes you have applications  or services that you want to script but you realise these applications themselves have their own terminals and their own  commands that work with their terminals only. You ask yourself what the hell? How am I going to script those inner applications’ specific commands and extract from these inner commands information to be used on the OS Level? Terminal multiplexers comes to the rescue. A good example  that I can give to elucidate this further is that of OpenSim, an open source multi-user virtual world server. OpenSim is a server that hosts and renders complex virtual worlds. It has its own commands and its own terminal/command line. I wrote an OpenSim – GNU Screen scripting blog article that I believe is very good to learn how to script OpenSim using a terminal multiplexer. I have used in the article the GNU Screen command.

2) Learn how to write scripts with multi-threading like effect so you can run things in parallel when you do some capturing of metrics etc. Please refer to an article I wrote on that.

The best way to learn Linux is by practice. You can start by installing a virtualisation software a.k.a hypervisor such as VirtualBox (Free) or VMWare on your host system, then install and run the Linux distribution you like and begin learning. It is quite easy. There is plenty of material on the web, YouTube, coursera, Udemy, Skillshare etc… It is also a good idea to learn how to export your own custom made linux distribution where you install your needed packages and then you export the whole OS as an ISO file or an OVA file (which works with VirtualBox/VMWare) so that you can use it in virtual machines or install it directly on your machine when you need this custom distribution.

To get you started, there is a good series of tutorials online for beginners :  Linux Survival (Module 1, Module 2 & Module 3). These are usually suggested to first year undergraduate students in the school of Computer Science at the University of St Andrews.

Great useful books on Linux that I have enjoyed reading cover to cover

There are many good online courses (MOOCs) on platforms such as Udacity, Coursera, Udemy, Skillshare etc…. There are also a lot of material on YouTube for free.

The following are few courses that I recommend on the Udemy platform. I am sure there are many others you can find:

Other instructors:

Practicing Linux terminal commands online:

Many tools allow you to write terminal commands and run them online from the comfort of your chair. These terminals are called Webminals or Web-based Shell Access Terminals. The following are few examples:

  • Webminal
  • Linuxzoo – “Learn Linux from the safety of your chair using a remote private Linux machine with root access”

Few tutorials and useful online resources:

Skill 2 – Master Python OR Perl OR MS Windows PowerShell

I think I have mentioned PowerShell previously for the MS Windows folks (the equivalent of Bash Shell scripting for Linux folks). It is used a lot by network and server administrators. If your PhD deals with Quality of Service (QoS) or system level performance measurements things of that sort. You need a good Language that helps you in your darkest days.

You might say, Ok!!! PowerShell for Windows, Shell scripting for Linux! that is too much! Well there is are many rings that rule them all, the most famous are mentioned below:

The first ring is Python: the king of all kings, the king of all languages, the jack of all trades, the master of all masters 🙂 since this language is extremely versatile and can do anything under the Sun (web development, scripting operating systems, machine learning, deep learning, data science and statistics, web scraping and list goes on and on till the end of time….). The language with batteries included.

Python is one of the easiest language to learn. Python is also widely used. Actually I have discovered in UK that they actually teach their students how to program in Python in their schools (i.e. at an A Level & even below that). This is just to let you know what type of environment you are coming to especially if you are coming from a 3rd world country. So a great number of students that start their undergraduate degrees in universities already knowing a bit of Python!

Python is an interpreted dynamically typed language. It is very readable almost like reading plain English language. It is very succinct with small code length. In other words, it is not verbose like the annoying verbosity you find in other programming languages. It is a multi-paradigm language supporting paradigms such as Object Oriented , Procedural, Imperative, Functional, Reactive among others. It is dubbed as the auspicious language with “all batteries included” in the package. Batteries included means that it has a big number of built-in modules out of the box and similarly a massive number of external & very efficient libraries covering anything you can ever imagine doing in programming.

Python allows you to write scripts exactly similar to what you can write using Linux/Unix shell scripting, in order to automate tasks of measurements or to capture metrics in experiments. The beauty of Python is that you can run the same script on Microsoft Windows, Mac OS and Linux or any other OS. Python is the language used  by default on your Raspberry Pi to program it to talk to sensors and actuators and is used  extensively in Internet of Things (IoT) applications and research.

Python and R are the most famous languages of Machine Learning and Deep learning these days. MATLAB is still a language that is used for machine learning but Python and R are the dominant these days. Python can also be used in your data analysis, in statistics and it allows you to draw the fanciest plots (data visualisation) through libraries such as seaborn, matplotlib, vincent among many others.

Python is used extensively in digital forensics, hacking and security PhDs. Probably this reminds you of Mr Robot series? executing Python backdoors and viruses from the terminal!!  Python is the most used language for creating easily hacking and security tools these days. A lot people are writing great hacking and security tools with Python including me (check my GitHub) and with utmost ease. This is the main difference from writing these tools with other ‘traditional’ languages. You can write  with libraries such as scapy , quite easily Network scanners, Man In the middle attacks (ARP Spoofers, DNS Spoofers etc…), Packet sniffers. You can write quite easily File Interceptors/File Injectors, JavaScript and HTML Code Injectors, credentials harvesters, enormous amount of malware (viruses, Trojan horses, backdoors…) and tools that protect against all these. Some of these tools are just literally few lines of code (I am not joking) and all the power is at your hand!!!

For aggressive hacking, all you need is an OS such as Kali Linux , [plus] mastery of Python [plus] mastery of Operating Systems and Computer Network Protocols and you can bring the biggest governments these days on their knees.

Black hat & grey hat hackers can buy from the dark & deep web many zero day viruses & python hacking tools undetectable by anti-viruses that actually work. You probably know that more than 90% of the web is hidden. They usually use a VPN software or the Invisible Internet Project anonymous network to avoid IP leakage and utilise powerful Linux distributions such as whonix, Tails, Qubes OS and Subgraph OS with a massive set of tools especially privacy-based web browsers such as the Tails darknet browser or Tor to access onion and clearnet websites.

Of course  always aim to be an ethical hacker and aim always to help private companies, institutions and governments with your knowledge if you are doing research into security  and hacking because black hat & grey hat hacking are morally wrong before being illegal. It is not different from robbing a bank or killing someone. Look at a concrete example, those despicable low scum bags who hacked the UK National Health Service (NHS),  and  by that hurting many patients who are old and vulnerable etc…! why? How inhuman and pure evil is this behaviour! So please understand that yes I want you to master Python but I do not advocate any unethical or illegal activities using it. I can understand if I try hard, people who hack the NSA or some corrupt governmental agency in the aim of exposing any wrong doing (i.e. whistle blowing), but I disrespect completely people who commit digital crimes and make people loose money or loose their lives.

So you can see again why if I asked in the past, myself with the knowledge that I have in the present, what language would help me a lot in a PhD, before even stepping into a PhD in Computer Science?  He will yell at me loud and clear, Python. Of course, if you are starting, you will learn Python 3. Actually this is irrelevant now since Python 2 will be dead very soon (thank goodness!!), no support at all for it but you might still be using Python 2 for legacy code so you need to understand the differences. Finally! this  nonsense of Python 2 and Python 3 is coming to an end!!

When I started my PhD, I did not knew Python, we never used it. I programmed in the past with languages such as C, C++, C# and VB (.NET),  JavaScript, Java and PHP but not Python 🙁 so when I came to UK and saw everybody using it, I was terrified!

Terminal Multiplexers in Python: Python can do terminal multiplexing also, recall when I talked about their importance and usage in the Linux skill section. You DO NOT need to leave Python toward shell scripting for this task, checkout as an example: pymux.

Disadvantage of Python: Python might be slow for some tasks especially if your doing some very complex tasks that requires high speed of execution or tasks that are very time sensitive. In this case, people usually use languages such as Rust or Google Go or C/C++ . Although, to be frank, a lot of python modules are optimised to run at C language speed. For more information, check these articles:  Why is Python so slow?, Why Python is Slow: Looking Under the Hood, and Yes, Python is Slow, and I Don’t Care. Python has also been criticised that it is a dynamically typed language, that it allows multiple inheritance which can cause problems,  that it does not have private attributes in classes equivalent to what the keyword private do in Java or other languages. There are only conventions in Python concerning what is private and what is not such as using the underscore. In addition, it does not have const: only conventions regarding that. Despite all these, Python is a very very beautiful & elegant language. It is here to stay and to grow.

Another language that is heavily used is Perl. Perl is also another “ring that rule them all” kind of language but in my opinion less than Python in prestige. It is ideal for scripting operating systems and is also cross-platform. One script for all of them.

I can see why network folks are attracted to it like a magnet. Perl is a very powerful language especially when it comes to parsing long texts for patterns. Perl supports object oriented, procedural and functional programming. Also a lot of folks are in love with the Perl language because of the good and many interfaces with C/C++ libraries. Perl used to be used extensively for server side web programming in the mid and late nineties but somehow died in favour of cooler kids such as Python, PHP, JavaScript and Ruby (especially when Ruby on Rails appeared).

Remember that scripting in this context, whether using Linux shell scripting, or Microsoft Windows PowerShell or using any other language such as Perl or Python, is essential in order to automate your experiments and to make them replicale by you or by other researchers in the future. In this major section, I am suggesting languages such as Python, Perl or PowerShell but it should be stated that any other programming language can be used for tasks’ automation or experiments automation. You can use PHP or JavaScript (via node js) or any other language for that matter. That has been said, languages such as Python, Perl, PowerShell, Unix/Linux shell scripting languages are very efficient when it comes to tasks automation and they contain many capabilities built-in and supplemented via additional modules or packages that makes your life a lot easier. In other words, these languages are commonly used for such purposes.

Master writing complex regular expressions. This is of course something that  you should already know by now. It is more of an undergraduate/junior level skill. Unfortunately, I have found that many PhD candidates struggle with this. I will give a concrete scenario of when regular expressions are needed. You are writing per instance Python scripts to capture traffic and different QoS networking parameters. You will probably use the Python subprocess module in order to call Linux commands (or whatever OS related commands for that matter) and you will receive back results or outputs. This output would be in the form normally of a long text with many goodies inside. So you need to sift and search all this text and retrieve exactly the particular small pieces that you want, so that you can use them in another process in your Python program or for whatever other purpose. In this situation you need to know how to write complex regular expressions.

There is an awesome website called Pythex that builds the regular expressions for you. It is also a good learning tool. You give it all the bulk text (normally the output of a Linux command that you want to extract or scrape data from and tell it what to do by choosing from a list of Regular Expressions cheat sheet. Then if you are using Python, you probably need to use an common Python module called re. Pythex website is not just for Python and can be used for any programming language.

Other useful online resources covering regular expressions

Please do not forget to document any script you write so that you and other researchers can understand them in the future. It is also a good idea to share them on GitHub or any similar platform so that other researchers can use them in their own experiments.

Few resources to just get you started with learning Perl & Python:

  • TutorialPoints – TutorialPoints is a great online website that provides guides to many technologies. They used to allow you to download the whole guide as a PDF now they require money and you can only download a small portion of the guide.  Search for Perl, they have a good introductory guide.
  • geeksforgeeks –  Another great website for learning major programming languages – This is just to get you started
  • Real Python Tutorials website
  • The Python lovely GitHub Cheat sheet – make sure you know everything here

Few Perl web resources – just to get you started – they are not really advanced:

Udemy courses on Python

There are tons of courses on Udemy covering Python and many of its applications in security, hacking, digital forensics, machine learning, deep learning, web programming, e-commerce, big data, data science and the list goes on and on…Some very few honorable mentions:

Udemy courses on Perl (beginner to intermediate level)

The Promise of Robotic Process Automation (RPA) for conducting automated experiments & tasks

RAP is amazing ! I remember back in the beginning of 2016, I read from cover to cover a famous book which is called “Automating the boring stuff with Python“. You can find it free to read online. I am sure a lot of Python folks  know it!!! I read chapter 18 which covered a cool Python library called PyAuto GUI. This library allows you to programmatically control the mouse and the keyboard in automation tasks. I have created couple of “software robots” using it. The functionality that I wanted was to fill out automatically very long online forms with specific data from MS Excel and then to submit them to a web server. The point that I am driving home here is that RPA technologies might be necessary for your CS/IT PhD especially if your research relies on conducting QoS/performance experiments or on automating cumbersome repetitive tasks. There are many tools that cost money such as Blue Prism, Automation Anywhere and UiPath  but many tools are now free and open source such as Automagica (Python Based), the famous Robot testing framework which does also RPA,  taskt which is a GUI based tool with no need for coding and the UiPath community edition (including UiPath Community Studio) which is a GUI RPA studio that is free but a bit limited. UiPath allows you to create very complex RPA workflows.

Creating RPA robots with UiPath Community edition is pretty powerful!! I am currently using it to automate some repetitive tasks in a project that I am doing for a client. The benefits of using a GUI studio or IDE-like tool for RPA is that you can achieve a lot of sophisticated automation with no code or with a very limited use of scripting. UiPath is only available on MS Windows (at the time of writing) and was written in VB.NET. It actually uses VB.NET as scripting language. It uses also XAML for workflow templates/projects. So you might need a bit of knowledge of the VB language to write VB functions and expressions. Don’t worry! the majority of the common automation tasks can be done via GUI i.e. via dragging and dropping different ready-made sophisticated automation activities.

Let us say you want to do an experiment measuring the performance of a certain tool that you have developed. You could write of course  Python, BASH or PowerShell scripts that automate things. During my PhD, I have written many scripts using these languages. Now with RPA available, free and mainly democratised, you can per example create an RPA robot using either a GUI IDE/Studio or non GUI scripting framework. This RPA robot will automatically open up your developed application, then it will monitor per example your developed tool’s CPU/GPU consumption using a benchmark application for per example 5 minutes capturing all performance metrics every 5 seconds and then it will close down all the opened tools & applications. All of this, is conducted while logging all the results into an MS Excel document in a very organised way i.e. with every metric value written where it should be written in Excel. The RPA robot can then send you an email when everything is finished. Add on that, you can make this robot do statistical analysis for you on the go. NB: In such a hypothetical scenario, you should not forget to have many baseline measurements i.e. measuring the consumption of the OS itself, the consumption of the monitoring application and that of the RPA robot so that you can figure out how much your developed tool is consuming CPU & GPU. I mean you can create very complex automations that are rule-based. You can even add UiPath machine learning activities and add-ons to your workflow to give your robots the capability to do judgement-based decisions. Check UiGo! which is a marketplace for all UiPath ready-made activities, snippets and workflows automation (both open source and commercial). To learn more about UiPath, please check the UiPath Community Forum.

Please consider learning at least one of the open source & free tools out there especially if your topic of research might need them! I should also mention  that testing frameworks in general have RPA or automation capabilities so learning them is also extremely beneficial especially if you don’t know one already. For example: Selenium WebDriver with Python which is designed to conduct extensive tests on web applications would also allow you to automates tasks.

Skill 3 – Master Version Control Systems & the Markdown language

Version Control a.k.a as revision control or source control, call them whatever you like, they are essential for any Computer Scientist or any software engineer or developer and Yes! even essential for a lonely wolf such as a PhD Candidate. Version control systems are used to manage source code repositories in the development process, they allow software engineers and developers to track modifications in the software and synchronise those modifications between a large group of developers. In addition, you can develop multiple versions of the same software from the same code base. Version control or Revision control systems are not only used for software development meaning they are not only used for your JavaScript, C, C++ projects, etc but they can also be applied to anything digital (yes!). They have been used in CAD (Computer Aided Design) systems, in 3D Models, in 3D Environments, in documents etc… The concept is there for everyone to use. Nevertheless, here I advise you to focus on the ones that deal with controlling and revisioning programming source code. Several famous kids in the neighborhood (there are others of course):

  1. Git version/revision control system
  2. Mercurial version/revision control system
  3. Apache Subversion
  4. GNU Project Bazaar

Please have GitHub and Bitbucket accounts

As a PhD candidate, it is important that you have accounts on both GitHub and Bitbucket for storing all your software that you have developed during your PhD unless if there is an embargo on them. Actually if you have been a programmer  and especially if you have worked in a company before the PhD, you would have got accounts on these services already, right? Even during the process of applying for a programming job in a company, these repositories constitute the portfolio of the projects that you have worked on or have contributed to and they showcase your abilities to prospective employers.

Furthermore, a lot of prospective PhD advisers now require their prospective PhD candidates to present Github portfolios of all programming projects done by candidates, in advance of moving forward with PhD research proposals.

Moreover, it is important from a PhD perspective, to reference all those software or tools that you have created during your research in your final thesis. This gives the opportunity for your examiners to look at the code of your projects, if it is deemed necessary. In addition, it gives everyone else the ability to use your tools. Isn’t one of the main requirements of a PhD, an original contribution to knowledge in the discipline? I hate when I see in academic papers or PhD theses, authors saying: we did this, or we did that! and then you say to yourself waw! that is awesome! so where we can we see this awesome project? Then sadly, you realise you can not find anything.

Now you might be working on a very sensitive piece of software that could be military grade or could be commercialised and you want to put an embargo on your work. A lot of these version control services offer you many techniques to make your repositories private or password protected. Of course if you plan on doing this, at the end of your degree meaning in a Viva Voce per example, you should give your examiners precise guidelines on how to access the repositories.

While on the topic, do you know that you can make your GitHub repositories citable? meaning that you can give them  Digital Object Identifiers (DOIs) so that they can be professionally cited. You might be asked to do that if you are submitting any programming code or datasets  as underpinning data or code when you submit your thesis. It does not harm to learn how to do that now. You need to create a Zenodo account and then to follow the tutorials presented in this article (Make Your Code Citable Using GitHub and Zenodo: A How-to Guide) or in this article (Making Your Code Citable).

Master writing in the Markdown language

Every repository you create for your PhD or for any other purpose for that matter, needs to have a very well-written documentation. Markdown is a text to HTML formatting language used by many programs and services such as GitHub, stackoverflow, Jupyter notebook among many others to facilitate the creation of fast beautiful HTML documentation.

I strongly advise you to learn Markdown. It is extremely easy. It won’t take from your time more than an hour to learn but it makes a big difference.

Start with reading the tutorials on GitHub: About Writing on Github articles. You can check the GitHub Flavored Markdown Spec  which is a long specification document that you do not need to read  but it is good for checking how to do in specific tasks in Markdown, then you can read Daring Fireball Markdown and the Markdown guide. There is a Markdown Cheat sheet which you can consult.

In addition, have a look at how famous GitHub repositories have written their own documentations.

Skill 4 – Don’t miss out on Computer Science Schools’ events/ Hackathons/ Game Jams/ Coding competitions, internships and opportunities to teach

Hackathons\coding competitions: Don’t miss-out on important Computer Science events and coding competitions. I know you are a PhD, you want to sit in your office. This is not good. A ‘Hackathon‘ is also called ‘Hack Day’ or ‘Hackfest’ and if it is dedicated for creating computer games it is called a ‘Game Jam‘. A hackathon and by extension a game jam is an event that spans a short period of time (usually 24 to 72 hours although could span few days) which involves participants usually programmers, games designers, software engineers etc. divided into teams that compete for prices. The aim is to produce a working prototype (not necessarily perfect) of a software, a web site, a game you name it in this short period of time (i.e. under the pressure of a deadline)

All prestigious schools of Computer Science in UK organise Hackathons and Game Jams every academic year. They are usually sponsored by big companies in the IT sector (Microsoft, Google, Facebook etc..). When I was doing my undergraduate, we did not have such luxuries and incentives! Students are very spoiled these days especially in the UK and the US! Hackathons are very good places for you to participate as a PhD candidate either directly in a team, or sometimes the School wants volunteers to help organising the event or supervising teams. Don’t miss out on these stuff, they are not only for undergraduates but also for postgraduates. Participation will improve both your technical and transferable skills such as problem solving skills, team work, presentation skills etc… Also this is good thing to put on your CV for both academic and non academic positions.

A lot of prestigious CS conferences have a form or another of a diluted hackathon. It is a recent trend! A lot of conferences have dedicated sessions that sometimes span the conference days and where researchers, from both industry and academia with accepted papers, can participate and collaborate on specific projects during certain sessions whether designing or implementing them. They have to present them at a specific day at the end of the conference. In many cases, this might be in the form of a full day workshop in which participants are divided into teams to work on particular small projects. You, as a PhD candidate, should be proactive and participate in those sessions when you attend conferences.

Internships: Some folks work in industry after the PhD others stay in academia. In other words, a PhD is not undertaken solely to help in acquiring a position in academia. In all cases, it is important and very rewarding if you can squeeze at least one internship during your PhD years if you can of course. Talk to your supervisor about that and try of course to find something that is linked directly to your own PhD research. In case, where an internship is difficult or isn’t an ideal choice, make sure to contribute to as many open source projects and work  on meaty programming projects in your non-PhD time. In your free time, build mobile apps, web apps or desktop software. Why am I saying this? Some folks aspire to work at Google or Amazon or any other big tech company after the PhD and these companies usually want to see a strong portfolio of programming projects or at least couple of good internships and industrial programming experience. To be frank with you they might not care about your research topic especially if it does not concern them directly. This is the case even for an industrial job that is not purely about software development.

Helping out in teaching labs, marking and tutoring: Help out the school in teaching, demonstrating, tutoring and marking Computer Science modules. Teaching a course obliges you to master it from my personal experience. “The best way to learn is to teach”. In addition, teaching experience is very good for your CV even if you do not want to apply for a purely academic position by the way.

Social media, interesting blogs & comics for PhD candidates

  1. Use Twitter: there are many twitter accounts that I strongly advice every PhD candidate to follow – check this link and this link. In particular I have found these to be very useful: Shit Academics Say: @AcademicsSay, @thesiswhisperer, @researchwhisper, @researchremix, @VivaSurvivors, @PhD2Published, and @scholarlykitchn
  2.  PhD candidates need comics to cope with the enormous stress and distress of the degree so a lot of them consult on a daily basis or some even print and put the most funny ones on their doors. The website is PhD comics. Another funny website is
  3. Few interesting and useful blogs: mine of course, the Thesis Whisperer blog and the Viva Survivors blog.
  4. The  Vitae website contain a wealth of interesting articles.

Other People good PhD skills recommendations

Computer Science Research and PhDs

  1. Recommendation by Toby Walsh – Computer Science and Engineering Department – UNSW Australia
  2. Recommendations by Prof Saleem Bhatti – School of Computer Science, University of St Andrews – UK
  3. Recommendations by Dr Tristan Henderson – School of Computer Science, University of St Andrews – UK

Other disciplines

  1. Recommendations by Prof Alex Wood – School of Psychological Sciences and Manchester Centre for Health Psychology, University of Manchester

Books to read

  1. How to write a thesis by Murray
  2. How To Get A Phd: A Handbook For Students And Their Supervisors by Estelle Phillips and Derek.S. Pugh
  3. The Unwritten Rules Of Phd Research by Marian Petre
  4. A manual for writers of research papers, theses, dissertations by Turabian

Take advantage of being a research student in UK

Don’t forget to take advantage of the following while you are student or in academia:

  1. Consider becoming a member of a professional body in your field if you are not a member yet. For example if you are a CS candidate, apply for a membership of British Computer Society (BCS). Membership of BCS depends on your experience.
  2. It is not a bad idea to ask at the start of the PhD the school or the supervisor for a new laptop for research purposes preferably a new MacBook. After 3 years or so, many schools ( not all of them of course), do not take back the laptop from you  since it would be old but the good news is that it will become yours. If what I said is not on the table, don’t forget that you can take advantage of the educational discount on Macbooks  from Apple stores. Gosh! I sound like I am making a advertisement for Apple here LOL  I know that Apple products are very expensive and not necessarily useful. It is like buying gold or some precious metals but there are legitimate reasons of why buy a Macbook and not any other laptop while being student. Well, first they are pretty decent high quality laptops. I mean I still have my Macbook Pro from 2010 and it is still functioning with great performance (touching wood) and with heavy duty usage. The Apple OS is awesome and is based on the Unix operating system. Second, if you are a computer scientist or a programmer you are bound to need a Mac at a point in time. I am a Linux guy but I believe there is a big need for a Mac PC or Mac laptop especially for developing MacOS, iOS, watchOS and tvOS apps. I could not find any other descent or legal way that really works around that. Tricks such as Hackintosh are illegal and actually do not work anyway. So a Mac might be needed out of necessity not necessary out of luxury. You can buy a Macbook Pro via an affordable installment plan (12 or 24 months). Finally, you can install MS Windows 10 on any Mac via Bootcamp  which is very cool. You get both worlds! You can also install Linux OSs via VM hypervisor such as VMWare or VirtualBox. I have a cheap separate Lenovo laptop with Kali Linux on it.
  3. Apply for a National Entitlement card – I have discussed this before in the article.
  4. Take advantage of Student beans, and of all the discounts and vouchers: discounts in restaurants and shops, discounts on software, even free software such as free Microsoft Office tools, free MS Windows educational licences, free IDEs etc. Discounts on buses, trains and airplane tickets.
  5. Take full advantage of your university careers center services: Interview simulators, many workshops, yearly careers fairs, CV/Cover letters private sessions, and massive preparation material for psychometric tests  – be aware that many UK universities have licenses for services such as AssessmentDayJob Test Prep or GraduatesFirst.
  6. Take full advantage of the hundreds of students’ societies
  7. Take full advantage of being on a university network i.e. with a university IP or having a university email. Why? Don’t forget that many academic services require registration from a university network (i.e a university IP) or a university email ending in You can not register with them otherwise. You can not even do that after graduation so do not forget to register with them while you are taking your degree. Examples: services such as or


[1] Müller-Bloch, Christoph, and Johann Kranz. “A framework for rigorously identifying research gaps in qualitative literature reviews.” (2015).

[2] The UK Quality Assurance Agency – The UK doctorate: a guide for current and prospective doctoral candidates found at

[3] Saunders, Mark, Lewis Philip, Thornhill Adrian. Research methods for business students, 7/e. Pearson Education, 2016.

[4] Fong, Philip WL. “Reading a computer science research paper.” ACM SIGCSE Bulletin 41.2 (2009): 138-140.

[5] Zobel, Justin. Writing for computer science. Springer, 2015.

1 thought on “Skills you must know before you step into a PhD”

  1. It was just a eye opener for me. The information you gathered here is definitely going to help me. Some advices about supervisors are very intimidating ,I hope that the one who helped me in making my research proposal for applying for Ph.D doesn’t turn out to what you’ve mentioned. Almost everything you mentioned here is difficult to access or try to learn as I don’t even own a laptop of my own till now due to some factor. I hope to save enough before even I clear my entrance exam so that I am aware how to use those programs which you’ve mentioned. I really hope this lockdown period gives me enough time to prepare well. Thank you for these valuable information.

Comments are closed.