Minimum requirements for a PhD thesis

I am getting tired of having to examine sub-standard PhD theses and then having to write the same old comments in the report on the thesis, so I have decided to set down the minimum requirements for a PhD thesis. Anyone who is contemplating asking me to examine a thesis should read these and comply with them or stop wasting my time. And be warned, you don’t want to get me cross by wasting my time with substandard theses that are an insult to my intelligence!

As most of those who read this blog are also academics, please feel free to add comments if you think I have been too harsh or lenient or when there are additional requirements that you would usually insist on. These requirements apply primarily to theses in molecular bacteriology and related disciplines and may not all apply in other disciplines.

The thesis as a whole

The thesis should comply with the regulations as regards margin sizes, order of front material and length of abstract. It should have been carefully proofread by someone experienced in the use of scientific English before submission.

The front material should include a declaration that all the work presented was performed by the student except where indicated and should list all the cases where the work presented in the thesis was not performed wholly by the student. Text from a multi-author paper, particularly if not primarily written by the student, should not be reproduced verbatim in the thesis without a clear description of where the student’s work ended and that of other authors began.

My suspicions will be aroused if the thesis is too short (<200 pages) or too long (>300 pages). It is almost never necessary to present a thesis that spans two volumes.

All sections of the thesis should comply with the usual conventions of scholarly scientific discourse. The thesis should be free of typographical, spelling, grammatical and stylistic errors. A consistent approach to typography (e.g. to fonts, use of capital letters in headings, justification of margins, line spacing and marking of paragraph breaks) and to formatting of references should be adopted throughout.

Latin binomials and gene names should be used correctly and italicised throughout, including in the table of contents and references. “E. coli” or any other species name should be treated as a singular noun. Proper use of nouns derived from Greek and Latin should be employed throughout, with care taken to avoid use of plural forms of the noun when the singular form is appropriate (e.g. do NOT write “the mitochondria is a bacteria”, “the culture media used was LB”). The terms “sequence homology” and “sequence similarity” should not be confused. Similarly, never say ORF when you mean protein-coding sequence or CDS!

Abbreviations should be spelt out at first use and defined in a list of abbreviations. Abbreviations should be avoided in the title of the thesis.

Throughout the thesis, when statements of scientific fact are made that have not been derived from the student’s own experimental work, they should be backed up by reference to peer-reviewed scientific publications. These may include review articles for subjects peripheral to the main thrust of the thesis, but should include original articles, with due credit for historical priority, for mainstream topics. References should be up-to-date and it should be clear that the student has kept up-to-date in the field right up until the submission date.

For the thesis as a whole and for each chapter, there should be a clear separation between the kind of material that belongs in the Introduction, in the Methods section, in the Results section and in the Discussion. There should be minimal repetition. In particular, the Introduction and Discussion sections of each results chapter should contain just enough reference to material introduced earlier in the thesis to maintain a narrative flow, but must not repetitively restate basic facts.

The Introduction

The Introduction should include extensive reference to the scientific literature, present a logical flow of ideas, with no non-sequiturs, and provide a rationale for the work that has been done. This need not consist solely of long stretches of text; instead, whenever appropriate, tables and figures should be used to convey the student’s understanding and knowledge of the field as succinctly as possible. The Introduction should be drafted in such a way as to provide evidence that the student had read and understood most of primary literature relevant to the thesis, rather than relying heavily on a handful of review articles. The student should also demonstrate an ability to make critical judgements on key issues. Irrelevant material should be excluded and students should from refrain from attempting encyclopaedic coverage.

In general, diagrams and figure legends should not be reproduced unchanged from papers and certainly not without the copyright holder’s permission. Instead these should be drafted from fresh or at least re-drafted to demonstrate that the student has grasped the concepts and facts illustrated or tabulated and has incorporated up-to-date findings and nomenclature.

The Introduction should close with a statement of aims and/or research questions waiting to be addressed at the outset of the project.

Methods and Results

Enough details should be provided of materials and methods so that someone expert in the field could reproduce the work—preferably without burrowing through a long chain of references Wherever possible, all data and analyses should be provided in the thesis, so that their veracity can be checked by the examiner. Where this might interrupt the narrative flow of the thesis, these should be presented in an appendix or, if datasets are large, they should be provided on a DVD.

Each line of experimental work should be given its own results chapter, with a brief introduction, a results section and a discussion section for each of the chapters. Methods specific to a given chapter might also feature in a dedicated methods section, although a Methods chapter dedicated to commonly used methods could be included.

The introduction to each results chapter should describe and justify the rationale for the programme of experimental work described in that chapter. Careful consideration should be given to what appears in the Introduction to the thesis as a whole and what should be reserved for the introduction to a specific results chapter, so as to present the most logical flow of ideas, while avoiding repetition. The narrative describing the progress of experiments and interpretation of data should deliver a simple and coherent description of the work undertaken. Long verbose descriptions of results should be avoided—instead, material should be summarised effectively in tables, figures or in crisp and lively prose.

Students should avoid over-interpretation of results—the student should distinguish between situations where the data are merely consistent with a given interpretation and where they actually provide convincing evidence for it.

Results from experimental controls should be presented whenever relevant. The number of replicates carried out for each experiment should be described and the reproducibility of techniques discussed. The potential for technical errors should be discussed together with the steps that the student took—or might have taken—to avoid them.

Where, for historical or financial reasons, the student has not adopted the most direct or most informative approach to a problem, this should be explained carefully. The student should attempt to evaluate the importance and shortcomings of each piece of work, in terms of how far it tells us new things about the world, represents an achievement in method development and/or provides resources for subsequent work.

Discussion

The thesis should conclude with a single Discussion chapter, which evaluates whether the body of research presented has fulfilled the aims set out at the end of the Introduction, examines how it fits into the broader scientific picture and assesses whether it is likely to stand the test of time in the light of recent technical and scientific developments. Repetition of material presented earlier in the thesis and a simple description of what has been done should be avoided.

Viva

During the viva, the candidate must demonstrate knowledge and understanding of all the work that has been done and the background to it. In addition, the student must demonstrate a sound grasp of basic concepts in microbiology and molecular biology (e.g. transcription, translation, the genetic code, Koch’s postulates), such as could be obtained from relevant undergraduate textbooks, and a critical approach to the definition of key terms (e.g. virulence, pathogenicity island).

Remember: the examiner’s word is final and there is no appeal against academic judgment, so get it right before you submit the thesis and certainly before you start the viva!


2 Responses

  1. herm
    herm
    August 10, 2011 at 3:05 pm |

    I can’t agree more. I’ve also been asked to examine a few of these theses, and you can find sub-standard theses by the hundreds on the online repositories of many universities.

    While reading your post, it occurred to me why tenured professors are so important, because only they could exercise their right to maintain academic standards with sufficient freedom. I have trouble enough with marking undergraduate exam papers.

    BTW, if you had reviewed any sub-standard material by means of me, please accept my apologies. There are many things that a non-tenured person cannot control …

  2. Morgan Price
    August 10, 2011 at 8:21 pm |

    I’m an active researcher in bacterial genomics but I virtually never read theses and I rarely see them cited. Probably they should be abolished. If that seems too radical, perhaps grad students should just staple 2-3 papers together with a detailed note on what parts they actually did (something like the author contributions that some journals have but a bit more detailed).

Leave a Reply

You must be logged in to post a comment.