Dissertation Tips from UH-ECON Thesis Advisors

The following are collected from faculty in the Department of Economics at the University of Hawaii.  Supervisors have different personalities, different work styles, and different economic methodologies.  Don't assume that the collected tips are true for all professors.

What key advice do you give all your thesis students?

• It is the responsibility of the Ph.D. candidate to show that he/she has made a contribution to the economics literature. That requires showing in what ways you have gone beyond the frontier of knowledge. Moreover, motivating your approach requires showing that other authors have solved related but not sufficiently transferable problems to fit the needs of your problem or they have developed models or methods in other contexts that are useful for your problem. All of these require discussion of the literature. But if you write a separate "literature review" section, it will not serve all these interests.  It is better to introduce discussions of literature where it belongs -- in the motivation, in the discussion of methodology, etc.
  Many proposals say that the student will do this and that (and then they end up doing something else). This is wasted effort. You should actually do something, e.g. get some theoretical results.  Then you can explain in the proposal what else you will do, and it will make more sense.
  Think results. Summarize your results in both the introduction and conclusion. (Don't worry about duplication.) How would you describe your results to an economist in a different field at a cocktail party? A non-economist? If you can't explain your results in simple terms, you don't understand them. [This is why John Kerry doesn't make sense (Stanley Fish, NY Times)].
• a.  Read...read...read;   b.  Schedule regular (e.g. every other week) meetings with your thesis advisor  so that there will always be a deadline to meet.
• Key means of communication is through written drafts.  Make it easy for the advisor to comment on them, i.e., your name, page numbers, double spaced,  adequate margins.  Important that students learn to express themselves in English.  Read The Elements of Style, Fourth Edition by William Strunk Jr., E.B. White, Roger Angell. 
• Pursue a topic about which your advisor is knowledgeable.  He will be more interested and will be able to help you more.
  
• The literature survey should be thorough. There was a case in the past in which the student said he did a thorough article survey but couldn't find any theory paper directly relevant to his dissertation topic. Out of my strong suspicion, a few months later, I went to the library and found a couple of directly relevant papers with an ease. Of course he was embarrassed, and I was very unhappy with him on that.
• Don't do a "review of literature." Discussion of relevant literature should arise more spontaneously in the context of particular objectives, especially motivation and justification of your methodology.
• Can your tell your grandmother (or grandfather), who does not have a Ph.D. in economics, what you are trying to say by writing a so-called dissertation? Come with a question that is written in a straightforward manner that anybody with a college degree can understand.
• Don't tailor your dissertation to your supervisor.  Find a supervisor whose style and interests suit yours.
• Students should share with the dissertation chair any problem they detect or come across over the entire course of dissertation rather than withholding it to themselves (until the chair eventually detects it) fearing that it may take a longer time to finish or it may invite a trouble. Students should know that academic integrity or honesty matters, and that the honesty expedites rather than jeopardizes the progress of dissertation.
• Keep their eyes on the prize. Completing a dissertation doesn't take brilliance, just perseverance. Set goals on when to complete various parts and work hard to meet those goals.
• Don't try to work for the whole committee. You work to satisfy your chair; he/she will tell you if you should consult another committee member about an issue.
• Do take initiative to perform research and to solve problems, so that your chair does not have to tell you the obvious.  But don't start off on a whole new track without consulting with your chair first--you can waste a lot of time that way.
• As for empirical study, a thorough theoretical background should be given as an absolute precursor to empirical analysis and there has to be a clear (not fuzzy or dubious) linkage between theoretical and empirical parts. The issue to be empirically resolved should be clear and specific.
• In preparing a dissertation proposal, it is very important to make it clear why and how the dissertation can make a significant contribution or bears a significant importance. If not, the proposal may risk being rejected.
• Do not write a proposal without at least one key result. (The result may be generated from a simple model/simulation, artificial data, etc. to be expanded, or it may be one of many results to come.)
• Less is more.
• Follow your passion.
• Prearrange regular (every two week, every month, whatever is comfortable with the chairperson and the student) meetings with the thesis chairperson. This tends to keep the student and the advisor on a timely schedule. This is my number one recommendation.
• Determine the objectives of the thesis and not lose sight of them unless a change is necessary! Then everyone needs to be consulted and brought up to date.
• As for empirical study, it is extremely important to know the availability of data. If it is an empirical paper, the proposal should include a discussion of data and econometric strategy, including anticipated problems and proposed solutions. Students should always assume there is a problem in the availability of data. In many cases, you may not be able to test a hypothesis which you think is interesting in part because there is no such information. Thorough literature survey can reduce the risk of setting non-testable hypotheses.
• Strongly encourage students to help each other out in learning how to use the computers and appropriate software. In particular, empirical study with survey data sets undoubtedly entail major time commitments in part because of software learning. Even for those students who have developed computer skills, this is really time consuming. Computer skills are just tools for efficient research. Help each other out.
• Try to replicate previous empirical results in which you are interested, on the condition that the research used a standard, easily available data set. Although the replication itself should not be an original research, this is extremely helpful experience in finding pros and cons of previous results, learning software, data manipulation, and so on.

What are basic etiquette tips that students should follow when communicating with their thesis chair and committee?

• Work closely with the advisor. Avoid using e-mail as much as possible.
• Keep a diary of your discussions and agreements with your dissertation committee chair.
• Keep your agreements.
• Always make and keep appointments with your chair.
• Always follow up a substantive meeting with a *brief* summary of what was agreed.
• Don't demand or expect quick turnaround on written materials that you want your chair to review. Most chairs will need a week or so to review a document.
• Always tell and show your chair what is new or revised in the documents you are submitting. Use sticky tabs of underlining or something to save your chair time. It will also mean a (much!) quicker turnaround.
• Make appointments with committee members in advance to talk about student thesis progress, and substantive matters.
• Please compare the previous empirical results with the revised one.  Faculty do not remember all numbers.
• Students must keep track of deadlines and strictly avoid inconvenient and short notice exams.  All oral exams should be scheduled well in advance and at the convenience of the faculty members.  Students should have the materials to the faculty well in advance and should follow up with the faculty to make sure there are no shortcomings.  If a thesis is based on one or two seminal articles, make sure that committee members have a copy of the papers.  Do whatever you can to make the job easy for faculty.
• Attach a copy of my previously edited version of your paper to the revised version (but photocopy one for yourself).  This makes it easier to check whether you satisfactorily addressed my concerns and helps me give you a quick turn-around time.

What are your pet peeves when working with thesis students?

•  a.  students disappear for months on end after a consultation;  b.  not follow instructions without good reason;  c.  show up at the last minute with a "thesis" and expect to defend in a few days;  d.  lousy work  and sloppy writing.
• As a committee member, however, my only pet peeve is students giving me a 200 page thesis a day or two before the exam. The most frequent substantive problem I've noticed is empirical conclusions that don't follow from the statistics, or estimation models that don't follow from the theory. However, if a student is smart enough, I won't be able to understand the thesis anyway.
  
• Students should not expect their advisor to be their technical assistant. I would be more than happy to work out a mathematical problem with them if necessary, but I certainly would not waste my time teaching them how to produce elegant equations or graphs.
• Students should keep whatever deadline mutually set. Otherwise, the expected progress of dissertation may well be seriously compromised to their disadvantage.
• Students showing up outside of office hours and without an appointment.
• Students missing meetings.
• Students submitting materials and expecting a response tomorrow.
• Students insisting that they are ready to defend when they are not.
• Laziness. If you know that there is something you should be able to do by yourself, then do it before coming to me.
• Students borrowing my books and keeping them for an extended period of time, or never returning them.
• Students who give you chapter 5 when they haven't determined what the objectives of the thesis (chapter 1) are.
• Students who disappear for months and then slip something in your box and expect you to read it in a week.
• A caution about trying to finish a dissertation long-distance.  They must not impose editing burdens on their advisor (even if the sucker is willing to take it on.)  They must expect that the format of the final doc will not be correct the first printing, and so they must figure out how they are going to handle corrections and submission.  And watch the damn deadlines!
• Don't expect an editor to fix more than spelling and minor grammatical mistakes. Clear thinking and clear writing are inseparable.
• Students who bring a journal paper and ask "I want to replicate the paper by using a different data set. Can it be a dissertation?"  (esp, the paper I know is very poor quality)
• Students who ask a ready-made data set from me.
• Students who leave town and don’t give me a copy of the finished product.
• Students should follow up on committee member’s comments after the proposal and the final defense.  I often spend hours giving careful comments, only to find they are completely ignored and never acknowledged by the student.  Students should personally and individually thank each of their members in an office visit after the oral exam is complete.  At that time, they should explain how they are responding to the members concerns (or justify not responding if that is appropriate).

What are some examples of best practices that have made working with thesis students a pleasure and joy?

• Once the student makes progress on an interesting problem, the student and advisor share the fruits of synergy and the joy of discovery.
• The occassional student is a real self-starter, and can make good use of the advisors time.  This student comes to each meeting with clear questions, and is able to show progress from meeting to meeting.  It is very rewarding to provide the gentle guidence to the motivated student.