Wednesday, December 28, 2011

How to examine a PhD thesis

In the UK and most European countries a PhD in a subject is the way to gain entry to serious academic discussions of that discipline. The structure of the PhD examination is designed to achieve this. The two criteria are:

(a) does the thesis make a contribution to knowledge?

(b) is the candidate a reliable member of the academic discipline?

As you can see, both the thesis and the person are being examined. Both aspects are crucial.

The understanding of what is a contribution to knowledge, across almost all disciplines, is that the candidate has conducted a systematic investigation into a question of interest to the academic discipline and has established a new (surprising) result. The words "systematic" and "established" imply that the thesis contains a rational defence of the result, which a sceptical member of the academic discipline will find convincing.

For the second criterion, about the person, examiners will start with the literature review that the thesis contains. The examiners will want to see that it includes references to all previous material that is directly relevant to establishing the new contribution, including papers that prompted the research and provided the framework or starting points, and papers that have reported results that appear relevant to the investigation. The way that these papers are reviewed is important: the PhD examiners will carefully consider the fairness and accuracy of the candidate's judgements on these previous contributions to the subject. Are any important contributions missing from the review? Are the candidate's judgements of the work of others well-founded?

In both cases, in the viva voce examination the examiners will probe any areas of doubt.

For the quality of the investigation and the reliability of the conclusions the examiners will want to check the research methodology and the coherence of the argument. A thesis is well-argued if all statements made in the thesis either are previously established knowledge or have been established by the candidate's investigation. There should not be repetition or non-sequiturs. The methodology and the breadth of the investigation will be satisfactory if they support the claims that are made for the generality of the conclusions. That is, even if the evidence that has been gathered has been correctly evaluated, what confidence can there be that further evidence would not yield different results?

For the adequacy of the candidate's knowledge and judgement, the examiners will carefully consider the list of references supplied and the way they are dealt with in the literature review. All bona fide members of the academic discipline will immediately notice any important omissions, especially of recent work relevant to the thesis. For the purposes of the thesis while it is important to acknowledge the first source of an idea, it is even more important to cite more recent work. And irrelevant material should not be included, as this would indicate that the candidate does not really understand the academic discipline.

So, in a typical case a PhD examiner will start with the research overview (typically in Chapter 1), will check whether a contribution to knowledge is claimed at all (first and last chapters) and if so, whether it really is a worthwhile contribution to knowledge. If the research turns out not to be in an area of interest to the examiner, or there is no contribution to knowledge, the process may well stop at this stage.

Next the examiner will consider the list of references, and then work in detail through the thesis, making careful notes. These notes will generally be of the following kinds:

(1) typographical or grammatical errors, or other inappropriate use of language

(2) factual errors about the academic discipline or its history, gaps in the reasoning or unsubstantiated assertions

(3) missing or irrelevant references or inappropriate comments on literature

(4) poor methodology or evaluation, or over-enthusiastic generalisation

Finally the examiner will check that the abstract is an accurate summary of the thesis, and return to the two criteria mentioned at the start of this article.