Like anyone carrying out an "assessment exercise", thesis assessors have certain questions/criteria in mind when carrying out their assessment. As found by Mullins and Kiley (2002), these questions/criteria typically include:

"How would they have tackled the stated problem?"

This is one way people work out if the approach was appropriate/effective.

"What questions would they like answers to?"

Since a thesis is a report of original research, most examiners start with the hope that they will learn something interesting, as well as being mindful that they are to assess the work as well. Consequently, given your stated topic, they formulate a series of questions to which they would like answers (which is what you also should do in both pursuing your research and in planning and writing your thesis), and obviously would feel a little disappointed if those answers weren't forthcoming.

"Do the conclusions follow on from the introduction?"

There are several reasons why the conclusions to a thesis might not follow on from the questions raised in the introduction. One is that in developing their methods, the student has focused more on the mere collection of data and not thought carefully enough about what sorts of data they will need to answer the questions posed in their introduction. Consequently, the conclusion relates to what questions could be answered by the data, not on how the data answered the initially posed questions. If this is the case, then this raises serious questions about the competence of the student to do focused research.

Another possibility is that because of various difficulties, the research ended up going in a slightly different direction to the one originally intended, which is not uncommon, or additional questions to those originally conceived came up, or the data ended up answering more questions than were originally conceived. One approach to dealing with these eventualities is to write things up as though you ended up where you intended to go all along. Framed in the right way, this is not being dishonest, for just as a literature review needs to be focused on the literature relevant to your thesis, not a review of all the literature you read during your entire candidature, so too a research article or thesis is generally a coherent story of how certain questions can be answered by certain data, not a story of all the dead ends and false trails you went down before you found the right path to solving your problem or answering your question. (In some cases though, there may be value in pointing out why a technique that one might think would be good to use ended up not being of value.) It might also be that you do some research which ends up tangential to the main thrust of your thesis. If such research can't be integrated into your thesis in a natural way, then it is perhaps best to leave it out and just publish it as a separate paper.

Another approach to dealing with having ended up somewhere different to where one initially intended to go is to explicitly acknowledge that this has happened and explain why it happened. However, this approach would probably only be used if where you ended up is only a little different to where you intended to go. In all cases, the thesis should present a coherent "story" (see Conceptualising what a thesis is as a whole).

"How well does the candidate explain what they are doing?"

Do you find it frustrating when you read an article and it doesn't explain why they did things one way rather than another or doesn't explain why they think a certain view is true? Well your thesis assessors will too, sometimes because they want to learn, but also because they want insight into your thinking processes so they can assess how good these are. And certainly subsequent students who read your thesis hoping for insights into how to think about things and why certain things were done or why certain things are believed to be true will find your thesis frustrating if their questions aren't answered.

Obviously one way to determine if your explanations are clear is to get someone else to read your thesis and answer this question. Another method is to re-read what you have written days or even a few weeks after you wrote it and ask yourself how clearly you have explained things. After a long break, you will have started to forget what you were thinking when you wrote something and you will read your work more like another person would.

"Is the bibliography up-to-date and substantial enough?"

"Are the results worthwhile?"

"How much work has actually been done?"

To obtain any research degree, your research must be of a sufficient size, scope and complexity, with a PhD being significantly larger in scope and complexity than an MPhil for example. So how can you tell when you've "done enough"? There is no easy answer to this question unfortunately, but your thesis advisor should be able to advise you, and perusing a number of successful theses from your School should also give you some feel for what is the required amount.

"What is the intellectual depth and rigour of the thesis?"

"Is this actually ‘research' - is there an argument?"