What to Expect

---

Thinking about a PhD in Accounting but not sure what it's all about or if it is the right choice for you?

• See the American Accounting Association (AAA) webpage "Thinking of a PhD?"
• See the article below by Waterloo PhD alumni Shane Dikolli.

---

By Shane Dikolli, PhD, School of Accounting and Finance, University of Waterloo.

I was once asked by an economist with a bachelor's degree, so, you are working on a PhD that specializes in management accounting - what does that mean? Are you able to add really big numbers together or something? In addition, many accounting PhDs and students I've met, including some highly respected accounting scholars, suggest they knew very little about what to expect when first entering a PhD program in accounting. The purpose of this article is twofold. First, it provides some insight into what one might expect as a student in the Waterloo PhD program in accounting. I don't necessarily think that these insights would be observable at all PhD programs in accounting, but from my experience in talking with people at dozens of programs across North America it wouldn't surprise me to find at least some of the insights being common to many different programs. Second, it provides a number of practical suggestions on how I went about eliminating obstacles I encountered during the dissertation phase of my program. These suggestions in particular potentially could apply to different PhD programs, different types of thesis-writing courses, and to PhDs in different disciplines.

Insights from inside the Accounting PhD program at Waterloo

I received plenty of warning before coming into the program that I should improve my math skills (specifically, calculus and linear algebra). Was that because as PhDs in accounting we really did have to add together very large numbers and needed extra math skill to be able to do it? No. Was that because Waterloo had such a strong Math faculty and every graduate had to have strong math skills? No. The answer is that a large component of accounting research relies on mathematical models that try to capture economic intuition and institutional knowledge about how accounting impacts the world.

This doesn't mean that all PhD theses must be based on mathematical modeling. The methodology chosen depends on the student's preferences and competencies. Instead of mathematical modeling, they can choose to develop predictions about what we might see in the real world, then either collect real world archival (or field) data, or, design and conduct a laboratory experiment, and finally test the validity of the predictions using (mostly) statistical techniques. Irrespective of the methodology chosen for the dissertation phase, the faculty emphasize that students should have some understanding all methodological types. Knowing math helps, especially in my case where I in fact completed a dissertation based on mathematical modeling.

The PhD in Accounting program at Waterloo and, from what I understand, accounting programs at many universities across North America, comprise two primary components: coursework and dissertation. At Waterloo, accounting PhD students typically study about 12 or 13 courses over two years. I was given suggestions for courses to do in the first and part of my second terms. As my interests unfolded, I chose the remaining courses that I thought would help me in the dissertation stage of the program. I completed four courses in the School of Accountancy, two in the Faculty of Mathematics, three in the Department of Management Sciences, and four in the Department of Economics. I think it would be ideal to try and develop a thesis dissertation topic during the completion of coursework. However, I found I was so busy keeping up with courses which were mostly outside my previous training (bachelors and postgraduate work in Accounting) that I barely spent the time thinking about a dissertation at this stage.

An important part of coursework over and above the requirements of each individual course are two components, assessable as pass or fail, and one non-assessable component. The first assessable component is that each student completes a Summer Research Project (SRP), conducted over the summer between the first and second year of his or her program. The results are presented to faculty usually in the fall following the summer the project is conducted. I could see that there could efficiencies, but more importantly a potentially stronger dissertation topic if the SRP overlapped the ultimate dissertation, however, this is not easy to do. My SRP overlapped my dissertation in the sense that they both used mathematical modelling, but the topics were vastly different. There is also the argument that having another working paper different to the thesis paper can help make a PhD graduate more marketable later on. However, my own sense is that given a choice between one paper that has extremely high credentials versus two papers that both have relatively lower credentials, the recruiters will be more interested in the one highly credentialed paper every time.

The second assessable component is comprehensive exams, held at the end of the summer between the second and third years. In total there's about 18 hours of exams held over four days and these can theoretically test anything that has been taught over the previous two years. The exams are usually open book, but who has time to look at notes when the examination hours seem to go so quickly? About six weeks into the first term of the program it is very easy to panic and wonder how it could be possible to ever sit comprehensive exams, but I know firsthand that the amount of knowledge that a student has just prior to writing comprehensive exams is awesome. It somehow comes together - a combination of hard work, commitment, experience, speaking the right lingo for two years, support from peers and faculty, and being curious about why things happen the way that they do.

The third non-assessable component is weekly research seminars. The School is fortunate in that it has a well-established reputation in the US, where many of the most highly published accounting researchers in the world are located. As a consequence, the School is able to attract a number of excellent speakers who always also spend time with the PhD students discussing all sorts of issues that may or may not be related to the paper that the speaker presents.

Once the comprehensive exams are done, the SRP is finalized, and all the courses have been completed, the coursework component is all over and there is a great sense of relief. Walking around campus knowing that there's no class to go to, no assignments to hand in, or no exams to study for, is all very liberating. But the relief is really only short-lived. A new set of stress-inducing choices needs to be made and different obstacles need to be overcome.

To get through the dissertation stage of the PhD program, the student must make three critical choices in addition to the choice of methodology (i.e. mathematical modeling, archival, field study or laboratory experiment). First, what theoretical discipline will drive the predictions that the thesis will make? Economics underpins much of accounting research. Research that focuses on the individual often use psychology or behavioural sciences as the underpinning theoretical discipline. Second, what functional area in the School of Accountancy is of interest? The school offers management accounting, financial accounting, tax and auditing (in no particular order) as functional areas of interest in which to base a topic. Third and most importantly, the topic itself must be pinned down. I think that topic selection is without question the most difficult and challenging part of a PhD dissertation. The topic must be interesting, original, rigorous, non-trivial, do-able and supervisable. It sounds daunting but the faculty are very skilled in knowing about topics that will fly. In addition, the training that a student receives during the first two years of coursework, I think, creates a different level of thinking that also helps formulate potentially viable research questions.

Dissertation obstacles

So what are the major obstacles during the dissertation stage? Many of them related in my case to topic selection and maintaining momentum in writing. If I had to start the dissertation over again, knowing what I know now, here's what I would tell myself:

I really know what you're currently going through. Topic selection is, without question in my mind, the most difficult and challenging part of a thesis. Once you have a reasonable idea of what the topic is, I think things do get easier. However, there's plenty more frustration ahead, particularly if you plan to do modeling. From my experience, I found it more efficient to identify the problem and ingredients of what I want to look at, determine intuitively what I think the answer will be, and then build the model. Early on I had this delusional idea that I would program all the ingredients into MAPLE software and the answers to "all the world's problems" would fall out. Wrong. I think it is much more productive to have a strong idea of what you think the answer to your problem should be and make sure that the model achieves it. Mathematical models that produce the thrill of discovery are extremely difficult to come by, or else my thinking is just not sophisticated enough to be able to program such models.

Here's seven things that I have done to help me keep moving forward:

1. Keep a binder of everything that you work on.

   Some suggested headings in the binder, which are borrowed from How to Complete and Survive a Doctoral Dissertation by David Sternberg (1981), New York, St Martin's Press:

   1. Topic selection (an ongoing journal of topic ideas)
   2. Devil's advocate (difficult questions and responses that someone might ask about your ideas)
   3. Troubleshooter (quick-fix solutions to problems found during, for example, model-building)
   4. How am I doing? (a weekly or daily journal of achievements)
   5. Timetable (lists key deadlines for getting things done)
   6. Serendipity/inspiration (a list of new ideas generated while writing the thesis - perhaps to follow-up post-thesis)
   7. Key literature (I don't keep these in the binder - I just keep a separate folder of the articles).

2. Keep a journal.

   Writing up either a daily or weekly journal of achievements is vital, for at least three reasons. First, there are no mid-terms, weekly classes or anything similar to check to see that progress is being made so a journal is critical to alleviate fears that you might not be moving forward. Second, writing also helps articulate your concerns more clearly - which often leads to more efficient problem-solving. Thirdly, the journal also keeps you up to date on where you are, in case you have to stop for a few days for whatever reason.

3. Just write it.

   This is a variation of Nike's "Just Do It".

   One of the turning points for me before my proposal was when my supervisor said to me that he couldn't properly assess my thinking/ideas unless I presented it to him in the form of a paper. I had been bouncing around out of control without a strong direction and finally his instruction got me focused. I didn't care that the wording might not have been exactly as I would have ideally wanted it, I just got it down on paper. It is then that you can exploit an outstanding supervisor's great strengths - he or she will usually provide quick, reliable, detailed feedback, which gets drafts of the chapter going. I was surprised how quickly things moved after I got that first draft of the proposal written.

4. Three-paragraph chapter ideas.

   Bill Kinney, a highly respected scholar at the University of Texas at Austin, suggested at the 1996 American Accounting Association Doctoral Consortium that if you can't summarize your paper in three paragraphs then the paper likely doesn't communicate your work well. The three paragraphs are based on the following questions: what is the research problem?; why is the problem important?; and what will you do to address the problem?

   This can form the basis of the introduction to a chapter. The second section in the chapter is the model description. The third section is the model analysis with extensions and you conclude in the fourth section. Works every time.

5. Get help on specific problems.

   I've noticed that whenever I'm really stuck, articulating the problem in writing and/or discussing it with a thesis committee member or fellow PhD student has generally helped tremendously. Sometimes the committee member or peer might not even be able to offer direct help but just discussing the problem out loud or simply having to articulate the problem invariably leads much more quickly to possible alternatives than "spinning wheels". Emphasis also on specific - the more specific your identification of the problem, the quicker you will find generate alternative solutions to the problem.

6. Commit to no distractions.

   I could quite happily surf the Internet all day, I reckon. But it does nothing for my thesis. I am very conscious of staying focused on what I'm doing and always asking myself (when I've allocated time to work on my thesis) that if what I'm doing will move my thesis forward.

7. "Class assignment context" work habits.

   When writing up the model, I try to work "as if" I am writing up an assignment. I impose an assignment deadline and meet it, just as I would if I were doing coursework.

Many of the above suggestions (e.g. journal of ideas, how am I doing entries, troubleshooter entries, devil's advocate entries) are operationalized as individual entries in my thesis log, which I've created using the journal feature in Microsoft Outlook. I print a hardcopy of each entry for my thesis binder. I realize that many of the suggestions might be more appropriate for once you have a topic, but I still think you can use the 3-paragraph method to pitch initial ideas to your supervisor and start writing up a "How am I doing?" journal each day. I now do a weekly journal (in the early going it was a daily journal), an edited copy of which I send to a friend of mine each Monday. My friend doesn't know all that much about my research, but he still sends me a rude note asking for the weekly journal report if he hasn't received it by Tuesday. It gives me an incentive to write that report.

This article contains a very small selection of experiences and lessons learned I learned along the way. They in no way necessarily represent the experiences of all students in the program and I have no doubt omitted experiences that other students would find particularly noteworthy. However, if this article provides just one piece of new information to a prospective student about what they might expect to find in an accounting PhD program, then it has served its purpose. I would be pleased to discuss anything I've raised with anyone interested in the PhD in Accounting at Waterloo.

Note: Shane Dikolli is now at University of Virginia.