The PhD Process

What the PhD Process is Like

We have talked about what a PhD is, and what is required of it. However, what is it like for the students themselves? What is the life of a PhD student really like? This brings me back to the time when I was an undergraduate law student. My friend introduced me to an acquaintance of his who happened to be an engineering PhD student. He was haggard and unshaved, and he looked as if he did not have any sleep the night before. He could barely stand up straight and was leaning against the wall. I was totally put off by that and told myself that I was not going to be a PhD student.

There is this common perception that a PhD student is very much like a prisoner locked up in an ivory tower. It can be true to some extent. The PhD process is in fact a very lonely one. I would say that being an introvert can be an advantage. Students interact mainly with the supervisor. Even then, meetings may take place at an average of once a month. There may be some interaction with other students, especially during the first year if there are taught courses. In the second year, there may be some interaction with respondents for students who collect data by survey. This is fairly common among those doing PhDs in organisational behaviour and marketing. Finance students usually make use of secondary data (data collected by other people for other purposes). Sometimes these data can be purchased from firms that do security analyses.

What is the PhD process like? The first meeting with the supervisor is usually as follows: The student informs the supervisor of his interests. The supervisor may agree with the student's proposal but usually will suggest some improvements. Some discussion may take place and a provisional proposal is agreed upon. Note that at this stage the proposal is only provisional.

The next stage is the literature review. A student begins by identifying the constructs such as personality, intelligence, work structure, job satisfaction, performance, culture, group behaviour, leadership, etc. that the student wishes to explore in his or her research. The student will go to the library and begin a literature search usually from computer databases using the relevant constructs as keywords. The databases will identify the titles and abstracts of all the relevant articles and books published. The student will have to photostat the relevant articles and borrow books. As a rough guide, a PhD student doing research in organisational behaviour or marketing will have to peruse about 1,000 articles and books. Out of these, the student will perhaps cite only 250.

Next comes the research methodology. Students will peruse through the same articles to see what instruments/questionnaires other researchers have used. They will try to identify which are the more popular instruments used by other researchers and evaluate the suitability of such instruments for their research. When you are at that stage, a valuable question that you should ask yourself is, ‘Are those questionnaires available?’ i.e. can you purchase them from the author? Another question that you may ask is, ‘Can these questionnaires be improved?’. This often results in you asking yourself (and your supervisor) the following series of questions:

1. Should I use the existing questionnaire in its original format?
2. Should I amend the existing questionnaire and use the amended version in my research?
3. Should I develop my own questionnaire from scratch?

My advice to you is, if you are able to find an existing questionnaire that is highly regarded by the academic and professional community and whose reliability and validity have been tested, you should use it. One example is the Job Descriptive Index (JDI) whose reliability and validity in measuring job satisfaction in a variety of occupations have been tested by numerous researchers (including myself), all over the world for over thirty years. (An instrument is valid if it measures what it purports to measure. It is reliable if the items are internally consistent with each other as well as over time). Similarly, should your research entails the measuring of personality, then it is best to use existing psychometric tests. Further details of psychometric tests can be found in Mental Measurements Yearbook. However, there may be several reasons why you may not wish to use existing questionnaires. One reason is that the objective of your thesis is to come up with another questionnaire that will overcome weaknesses inherent in existing questionnaires. Another reason is an economic one. Use of existing questionnaires often involves purchasing the questionnaires, answer sheets and scoring keys from the author. If your sample size is over 1,000 and you intend to use three or four different types of questionnaires, the costs can be huge (to a student undertaking his PhD without sponsor). The third alternative, which is to develop your own questionnaire, should be avoided unless developing a questionnaire is the objective of your research. This is because it is very difficult to defend the validity and reliability of such an instrument in the viva.

A full-time PhD student should commence with data collection during his or her second year. If the research method is a survey questionnaire, the entire process of data collection and coding can take anything from four months to a year. This is because the respondents/companies are only willing to be interviewed at times that are convenient to them. Thus, there will be some idle periods when you will not be collecting any data. In my PhD research for example, the pattern was usually for me to spend perhaps a week at one factory testing the workers who were available there, and then to return the following month to test some more workers. As the companies that participated in the survey were keen on minimising disruption to work flow, they advised me that testing should be staggered over several months. Workers who were not available on the earlier session would have an opportunity to participate in the subsequent sessions. Visits were always arranged in co-operation with managements of the various companies in order for them to be made at times when it would be most convenient for them to test their workers.

The next stage is data analysis. Should you do a qualitative or quantitative analysis? A quantitative analysis usually requires the use of a computer software such as the SPSS, Excel or Econometric Views. Qualitative analysis, on the other hand, may require the use of computer software such as NUD.IST or no computer software at all. My recommendation to students wishing to do research in organisational behaviour or marketing is to opt for quantitative research. In other words, my motto is, do quantitative research if possible. Having said that, there may be some areas in the social sciences where quantitative analyses are rare such as those involving corporate strategy and international business management. Here, statistics used are nominal—usually frequencies, percentages and means. Quantitative and qualitative research are not mutually exclusive. Sometimes findings in quantitative research can be supported by findings obtained by qualitative means.

If you opt for quantitative analysis, you will spend many hours in front of the computer. Some might find it tedious and boring. However, this process enthrals me and I find that I can spend weeks sitting in front of the computer from morning to night, experimenting with different types of analyses. Time really flies! The initial stages of a quantitative analysis are as follows:

1. Inputting data.
2. Preliminary analyses—checking for errors, missing cases, normality of distribution.
3. Conducting statistical tests like correlation, regression, factor analyses.

My description of the PhD process seems to indicate that it is sequential, i.e. a student does the background work and writes Chapter 1, then does the literature search and writes Chapter 2, and so on and so forth. However in reality, the process may be nothing like that at all! For instance, I wrote the literature review simultaneously with the research methodology. My introduction chapter was written simultaneously with the conclusion chapter, after the analyses were conducted and written out. When I was in my first

year, a third-year student once advised me, ‘Finish off your literature review in the first year. Get it over with and concentrate on the data collection, analyses and writing the other chapters in the second and third years.’ I tried to take his advice, but it did not work. I found myself going back and forth, constantly revising and amending what I have already written in all of the chapters. I was even amending my literature review about one month before I submitted my thesis. Anyway, my friend who advised me was still doing his PhD when I attended my convocation in July 2001. I heard more recently that he had sat for his viva and was required to make major corrections to his thesis. Three years later, he had still not managed to make the corrections and he subsequently dropped out of the programme without managing to complete his thesis. It is clear that his strategy did not work!

When to Start Writing

The questions that one should logically ask are:

• When should I start writing?
• Should I start writing as I go along, or after all the data have been collected and analysed?
• Should I write the chapters in sequence?
• Should I write first in longhand or type straight into the PC?

I would recommend you to commence writing as early as possible and not leave it all to the end—that is the advice I give to all students, even those at the University of Malaya who have to attend classes and examinations as part of their coursework. Writing a thesis involves a great deal of work and is often grossly underestimated by those who have not written a thesis before. Some students, during their second year of the PhD course, have advocated, ‘There is no point in writing too early as we may change the direction of our research as we progress. Thus all that we have written earlier has to be thrown out—a waste of time. After finishing my analyses, I am going to spend the next three months, solid, in writing up the whole of my thesis.’ Students may make the above remarks because they are motivated to conserve effort. They do not wish to do any work that might be thrown out later. In other words, they want to minimise ‘wastage’ of time and effort. However, they do not realise that writing a thesis can be an extremely time-consuming process. Furthermore, it is only through writing and rewriting that we can discover flaws in our arguments. Yes, I said rewriting. In fact, a thesis has to be rewritten several times before it approximates perfection. Rewriting a thesis or ‘polishing’ as some students refer to it, can take as long as one year! Why does it take so long? Well, as I mentioned in earlier chapters, a thesis is a position that you wish to argue. As such, your arguments must not contradict one another. You have to literally be able to defend every statement that you make. Rest assured that if you do not identify and eliminate contradicting statements, your examiners will, and you will automatically reduce your chances of succeeding in the viva. I have often observed my MBA students simply copy and past sentences from their literature searches. I can easily spot when a student has done this. The result is often an obvious lack of continuity and flow. There are often disjointed themes and sentences which do not make any sense. I have often asked the students to explain to me any ambiguous sentences. I find that they are unable to do so, for they have merely lifted portions of text from the database without fully comprehending its meaning and relevance to the current research. This exercise would be fatal if it were done by a PhD student.

Regarding the third question, I believe that the literature review should be written first. Start writing as soon as you have collected and read about 100 articles. At this stage your draft will probably look like a mere recitation of cases, with the very occasional criticism here and there. It is quite all right at this stage. Writing out the literature review early helps you to organise and categorise the articles on paper as well as in your mind. As you read further, more articles can be added and old ones rearranged. You will begin to see some patterns and will be one step closer to being able to make a critical review of these studies. As for the research methodology, start writing as early as possible. You do not have to wait until you have finished writing your literature review. These two chapters can be done simultaneously. Start writing on the descriptions of the instruments once you have decided to use them. Write about how you selected your sample and describe how you collected the data, as soon as you have commenced. The reason for doing so is that it is easiest to write about how you have gone about collecting data when you have recently done it and while it is still fresh in your mind. Results should be written shortly after they are obtained. The results chapters should also be amended from time to time as new analyses are conducted. This is because you may probably think of conducting further analyses, after you have obtained results on your earlier ones. The introduction and conclusion chapters should be written last. In short, a thesis is rarely written in the order that it is presented.

The final question is whether to write first in longhand or type straight into the PC. A classmate of mine preferred to write his first draft in longhand, amend the same a few times, and then type it in a computer. This was probably because he did not have his own personal computer, and English was not his first language. On the other hand, you may still prefer to write the first draft longhand, even though you have your own computer and English is your first language. This may be so because you find it easier to write it all down first and then look critically at what you have written. Basically, whether it is better to write the first draft in longhand or type it straight into your PC depends wholly on the individual. I have heard of a professor who insists that all his students write their first draft in longhand. I feel that it is wrong to make all of your students do so just because it works for you. When doing my PhD, I typed all my drafts straight into the computer. However, when writing this book, I wrote my first draft longhand. So, it appears that there are no set rules, even for the individual. If there is a rule of thumb, then that rule must be ‘Experiment with different methods, and do whatever works for you at the time.’

The Importance of Having Multiple Back-up Copies of Your Work

It is critical to not only save your work but to have multiple back-up copies stored in various media in various locations. During my second year, my house was broken into and my computer was stolen. I lost all the valuable data and research in my hard disk—amounting to two years' work. Fortunately, I had backed up most of my data in an Iomega Zip drive. As the last time I saved my data was a week earlier, I lost only a week's amount of work instead of two years' worth. I would also recommend that you save data in different formats and in different locations. For example, you may store your entire data on an external hard drive (this is highly recommended and it is what I do now) and keep this hard drive in your house and store another set into CD-ROMs in your office. So if there is a fire in either your house or your office, your data will still be safe. It is highly unlikely that both your house and office would catch fire at the same time (unless your house is your office). It is important not to keep your computer and back-up copies in the same location—and especially not in your car! While I was in Manchester, I saw a note written by a very desperate PhD student. It went something like this: ‘Please, whoever who took my laptop last week from my car, please return only my data as I am a PhD student in my third year and about to complete my thesis. You can keep the laptop—just return my data and I promise not to press any charges.’ He had kept all the hard copies and the soft copies in the same briefcase—a rather foolish thing to do. I keep two back-up copies—one in the office and one at home. The reason is that I sometimes keep my computer at home, and sometimes in my office. Having both the computer and the back-up copy in the same location is not wise. Another factor which I had to worry about (but you do not) was that my PhD was conducted during the infamous year 2000. I not only kept the soft copies in multiple back-up copies but I also printed all of my data and drafts in preparation for 1 January 2000. I spent New Year's Eve not partying with friends, but at home in front of my laptop watching the clock tick past 12 midnight and watching with great apprehension, my data and drafts—that's the commitment and passion that I had for my PhD! Fortunately, the Y2K problem was greatly exaggerated and I experienced no problems.

Managing Your Relationship With Your Supervisor

What is the role of the supervisor? What are the supervisor's duties? A supervisor is not a coach. A PhD student is not supposed to be instructed on what to do, every step of the way. This is one of the core differences between a PhD student and an undergraduate or MBA student doing a thesis or project paper. A supervisor's duty is merely to lay down the boundaries in which a research is to be conducted. Thereafter, the student is left on his own. Of course the supervisor gives help and suggestions but only occasionally. The research has to be the student's own and not the supervisor's. A PhD is in reality a training ground for a person to be a researcher, not a research assistant. The student must learn to exercise independent thoughts. My external examiner once told me in words to this effect: ‘A PhD is not just about writing a thesis. It is the process of self-discovery, adventure and agony. The pains and disappointments of the process of exploration have to be experienced by the student first hand. This process builds character.’ This was certainly what I encountered when I was a PhD student.

I have seen some supervisors sitting next to their students in front of a computer, teaching their students how to do analyses. I have also heard of supervisors correcting the students' first drafts, making amendments and even writing out whole paragraphs for them. Incidentally, I have also told my supervisor of this, in the hope that he will do the same. However, my supervisor said that what they (the other supervisors) did was wrong. In other words, my supervisor refused to do the same. Although I was a bit disappointed at that time, I realised subsequently that he was doing the right thing. In fact, one of the common questions that examiners ask the student at the beginning of the viva is ‘To what extent is this thesis your own? Who came up with the idea?’

A student who receives instructions from the supervisor every step of the way may love it and be able to finish his PhD within three years. However, the student has short-changed himself. He has not gone through the process of being lost in exploration and experiencing the subsequent joy of discovery. He has deprived himself of valuable training and experience and is not developing his full potential as an explorer and inventor. A PhD student should be an explorer, crossing uncharted waters, and not a passenger in a taxi. Unfortunately many people I know wish to be taxi passengers. They just want to take the simplest and quickest route to get a PhD.

My experience as a PhD student in the UK has taught me that generally supervisors expect:

• their students to be independent
• their students to be excited about their work, able to surprise them and fun to be with
• their students, when producing written work, to have their ideas reasonably thought out, typed and checked, rather than a roughly sketched first draft
• to have meetings with their research students only when necessary and that students should not waste the supervisor's time
• their research students to be honest when reporting on their progress
• their students to follow the advice that they give, when it has been given at the request of the student.

My Experience as a PhD Supervisor

• My experience as a PhD supervisor, on the other hand, has resulted in me laying down certain rules for my students. Some of these rules are also adopted by my colleagues. These are essentially my rules and I am not saying that all supervisors must adopt them. It is entirely up to them and the students. My students who do not like these rules can always find another supervisor. My experience is that it is important to clarify the rules from the beginning as it helps to manage customers' li>Students must not expect me to proofread their drafts and to go through them line by line correcting mistakes. They will have to check their own work. I recalled my supervisor telling me, ‘There is something wrong with your draft and I will not tell you what it is—you have to find it out for yourself.’ Initially, I found this extremely hard. However, after a while, I became competent in being my own supervisor. The experience was invaluable and I became a different person—from a demanding spoon-fed person to an independent critical researcher—thank you, Colin!
• Students should meet me without fail at least once a term and not more than three times (therefore twice a term is optimal). This will ensure that students will be motivated to make continuous progress. Often, students (not mine as yet) will not see the supervisor for several months, and when they finally are making progress, they want the supervisor to drop everything and interact with them every other day! This wreaks havoc on the supervisor's time management. Students have to remember that in most universities, supervisors are not paid an allowance nor have their teaching workload reduced because they supervise PhD students. They are the ones who need the supervisor and not the other way round. It is not my intention to upset you readers out there. Because I care for you, I am being perfectly honest and saying it the way it is.
• The supervisor can (not must) lay down the boundaries in which the research should be conducted. For example, the supervisor can instruct the student to conduct the research from the viewpoint of person-environment fit theory, strategic fit or culture. This is appropriate where the supervisor is well-versed in the area and knows that valuable contribution to the literature can still be made. Also, if the student decides to adopt a different approach with which the supervisor is not familiar with, the student will be on his or her own and the supervisor cannot guide the student effectively. Having said that, I know some supervisors actually prefer their students to go into areas with which they themselves are not familiar with. This is done so that the supervisors can broaden their own knowledge.

The above are my rules and my rules alone. After all, it is the student's thesis, not the supervisor's. It is common among students from the Far East, especially those who are fortunate enough to have parents who can afford to send them to private colleges, to expect teachers to spoon-feed them with notes. It is not the duty of supervisors to spoon-feed their PhD students. The time has come for such students to learn to think for themselves.

I mentioned in my earlier book that supervisors expect their students to be excited about their work. After all, no one is forcing them to do a PhD. Students must also be able to surprise their supervisors by showing them some material that their supervisors have previously not been aware of. Now you might say that is impossible. If the supervisor is supposed to be an expert in the area, how can he be ignorant of something that the student was able to discover? However, this is not only possible, but occurs more frequently than you think. In fact, I was able to do just that during my third year. It is not a sign that your supervisor is incompetent. Rather it is a sign that you have made real progress. During the first year, the supervisor should know more than you regarding the research area. However, as the research becomes more focused, the student should become more of an expert than the supervisor, at least as far as the research problem that the student has been working on is concerned. This is because, by the third year, the student has been working on the problem far longer than the supervisor has. However, I fully acknowledge that there may be some supervisors out there who are so far ahead in their fields that students may never catch up with them.

It is not always necessary to show written work to the supervisor. Usually, I only produced written work for my supervisor when I felt that such a method of communication was clearer than merely explaining to him orally what I had found out or the line of argument that I wished to take. However, whenever I did so, I had it reasonably prepared, typed and checked, rather than a roughly sketched first draft. It is a matter of courtesy to spend some time on the material, if you expect your supervisor to do the same.

Supervisors expect to have meetings with their research students only when necessary and students should not waste their supervisors' time. I met with my supervisor only when I felt that I had something important to discuss. Meetings were more frequent in the first year—about once a fortnight. Meetings were less often during the second half of the second year when I was collecting my data. Meetings were even less often during the third year because this was the write-up stage. However, I did send him a lot of written work at that time. There were a few rare occasions that we met more than once a week. This may contradict my earlier rule about frequency of contacts. However in my case, I was his only PhD student—so my supervisor had lots of time to meet up with me (which he'd probably disagree with). I also remembered a time when I did not see my supervisor for six months. This was because I was preoccupied with domestic issues at the time and my work progress was rather slow (at first our house was broken into, then we moved house, and finally my wife gave birth to my beautiful baby daughter). Thus over the three-year period, I met with him for an average of once a month. A general guideline that we agreed upon from the start of the course was to meet as often as we really needed to. My classmate had a supervisor who insisted that they met once a week. He would work into the early hours of the morning before each meeting so that he would have something to show his supervisor. Needless to say, this caused him a lot of stress. Subsequently, the supervisor's other workload increased and he was no longer able to keep up with the weekly meetings. They subsequently abandoned the idea of meeting once a week and met only when necessary, which was approximately once a month.

Students should always be honest with their supervisors when reporting on their progress. There is no point in lying to their supervisors; they would only be deceiving themselves. If the student is behind schedule for any reason, then it would only mean that the PhD would take longer to complete. This is a fact of life for a PhD student. More often than not, students end up taking longer than they initially planned at the start. It is not the end of the world. In fact, further determination and perseverance, rather than despair, is required in such situation. I know of one person who boldly proclaimed that she was going to complete her PhD within three years; at the fourth year she still has not completed her PhD. Another classmate of mine also told me during our second year of the PhD candidature that he would take a few more months to complete his thesis, i.e. well within three years. I submitted my thesis in the third year but he submitted his only during his fifth year. So much for ambitious proclamations!

At the University of Malaya, we have a Doctoral Colloquium in which students are required to present twice a year (i.e. once per semester) the current status of their research. The Board, consisting of the chairman, the supervisor and another lecturer, will listen to the student's presentation for half an hour. The Board can make one of the following recommendations:

1. Allow the student to continue with his or her research, or
2. Allow the student to continue with his or her research but with close supervision because some but insufficient progress has been made, or
3. Terminate the students, candidature on the basis that the student's performance has been unsatisfactory for at least two semesters.

Finally, supervisors expect their students to follow the advice that they give, when it has been given at their request. I have observed some of my colleagues (PhD students) asking their supervisors for advice, and if the advice given was not what they (the students) wanted to hear, they ignored it. I feel that such behaviour is very insulting. I can relate that to my experience as a practising lawyer when I was often sought for advice by my clients. When the advice that I gave them was not what they wanted to hear, they would ignore me. Worse still, they would even try to convince me to give them the advice they wanted to hear. I can therefore sympathise with a supervisor who has faced similar situations in the past. So, please, do not do the same to your supervisors. I know of supervisors who have actually resigned (as supervisors but not as academic staff) as a result of this. They claimed that they have requested their students to make specific changes and that the students have refused to do so. Although I have said earlier that students should exercise independence, this extreme form of independence is highly inappropriate.