theEzone

Description versus interpretation

A recent examiner has noted that the thesis he was considering was heavy on description but light on interpretation. He thought that descriptive theses are acceptable at the Masters level but that the goal of the PhD is to generate new knowledge and go beyond description.

This is an interesting point that has come up in lots of examiners' reports. It is at the core of the question about how much theory a PhD thesis should contain. So what is the answer?

Our university's definitions state the following:

The Masters thesis must be a substantial work generally based on independent research, showing sound knowledge of the subject, some evidence of independent thought, and it must be presented in clear and concise language.

The PhD thesis must make a substantial and original contribution to knowledge.

Neither of these definitions makes reference to a distinction between description and interpretation, nor to theory, yet this is a point that examiners come back to again and again. A thesis that merely describes the state of the subject area (the facts) will have more difficulty passing as a PhD than one that explores the meaning of these facts. The meaning, or the interpretation, is the basis of the theory, and a theory has predictive power: it allows the consumer of this new knowledge to know more than the mere collection of presented facts. A theory provides an insight that can be applied to other, similar, systems of study.

theEzone is still alive

It's been a while since I've blogged on theEzone, but not for want of having something to say. Rather, time has been the issue.

However, I thought a short note on where we are up to so far this year might spur on a few near-completers.

Since January 1st this year there have been 114 completions lodged with the Graduate Research School, and 10 have been awarded with Distinction. A completion is defined when the examined, corrected, final version of the thesis is lodged with the GRS and the Library. Some of these completions will also have been examined and classified this year, but some will be the result of examination last year.

So far this year, the Board of the Graduate Research School has classified 94 theses, with a further 11 under review at the moment. A thesis is deemed classified when all the examiners' reports are in and a decision on the fate of the thesis has been made. A typical classification would be "Pass with minor revisions, carried out to the satisfaction of the supervisors and the Head of School". An initial classification is made by one Board member, the nominated Recommender, and then the rest of the Board "comment" on this classification by either agreeing or not. Typically, the classification phase takes about 2 weeks after all the examiners' reports are in.

In addition to this, there are currently 163 theses submitted for examination. These are all in the process of either being sent to examiners, being examined by examiners, or having their reports chased up by our examination officer.

Last year we completed 306 theses in the calendar year: a record. At this time last year we had completed about 80, so we are well ahead at the moment. Submissions come in fairly regularly over the whole year, although there seems to be a little blip of activity in completions over the June - July period. Perhaps this is all those candidates who want to attend the September Graduation ceremonies?

 

The reflective examiner

I love it when an examiner talks about "the PhD" in general, rather than simply the particulars of the one under consideration. When this happens, we get an insight into what examiners are looking for. Usually, such reflection occurs when the examiner is trying to explain to the candidate how a thesis might be improved - this is the formative part of the report.

Here's a summary of the sorts of comments that often arise:

Normally in academic writing, especially at the PhD level, it is crucial to begin with a statement of a research problem. A research problem is at the core of an academic study. It is the main idea of the research and represents the philosophical view of the PhD candidate. Without this philosophical view, the thesis only deals with technical aspects and this does not suffice for a PhD.

A research problem is created by intensive, extensive, comprehensive, and critical reading of the literature. Any gap in the literature, caused by methodological problems or incorrect interpretations of data, can become a general statement that can subsequently be enhanced by the researcher's opinion on the issue. Thus, there must be old knowledge and new knowledge in the thesis.

A literature review is not a re-echoing of the literature; it must be a critical review of it. First and foremost, a literature review is the researcher's own opinions and comments about previous writings and studies in the area. These opinions are built from the reading. They may be directed towards methodological shortcomings or weaknesses in previously published work. These opinions of the candidate should be categorized, labelled as sub-topics, and discussed. Additionally, some part of the literature review will be a gathering together of known data, statistics, opinions and definitions that will be used to justify or support the argument put forward by the candidate.

Images of originality

Any diagrams, figures, or images in your thesis must be treated in the same way as your text: they must be your own, or they must be appropriately cited. However, unlike text, when an image is copyrighted (and it is safe to assume that all are unless it is explicitly stated that they can be used with a citation only), then you must obtain permission to use someone else's image.

The over-use of figures downloaded from the internet detracts from the quality and it is preferable to see the candidate's own images, which then would be clearer and in some instances more appropriately targeted to the relevant information.

Justifying your thesis

I know it is hard enough actually stating what your thesis is, but once you have articulated it, you then have to be able to convince the world that it is important enough to be worthy of 3 years' consideration, and that what you have found is a substantial contribution to the world's knowledge.

One of the most serious concerns relates to the need for a sound justification that the research topic is substantive enough to satisfy the requirement that the thesis makes a substantial and original contribution to scholarship.

One way of making this justification is to demonstrate, through your literature review, that other people have identified the gap in knowledge that you are intending to fill, that they care about knowing the answer, and that the question is contemporary and relevant. In this way, citation becomes a powerful ally in your construction of a sound argument.

Me Write Pretty One Day

Examiners all appreciate a good story.

The entire study was characteristic of being well thought out, well organized and professionally and scientifically conducted. The Thesis (and research) flowed very nicely from one step to the next and told a very nice, needed, related story.

A good story needs a number of ingredients, including real characters, mystery, rhythm, questions, anticipation, surprise, depth, contrast, climax, and The End. While this advice comes from science fiction story telling, it can also apply to telling a good research story.

To add to the argument, I have posted an article called "Me Write Pretty One Day: How to Write a Good Scientific Paper" by William A. Wells. 

A good idea

Sometimes, in passing, examiners present good ideas for all.

The abstract should have a short paragraph in which contribution(s) to knowledge are claimed and the major finds of the thesis and their implications noted ... It would also help retrieval of the thesis by interested parties, if a series of keywords are added to the bottom of the abstract.

Text, especially digital, should  be tagged by keywords and appropriately structured to facilitate digital search. Where appropriate, accompanying data should similarly be tagged and structured so that direct links can  be made between the text and the data. Simple, old-fashioned, examples of this idea include putting your text within chapters and chapter subsections, identifying and tagging citations, equations and diagrams, and forming links between these objects and the text itself. Pushing this idea further, if your thesis is accompanied by the data you collected, you can link your text, diagrams and tables etc to the appropriately tagged component of your full dataset.

And speaking of digital theses, I'm wondering  how long I have to wait before someone requests permission to submit a purely digital thesis with embedded objects such as sound files (interviews, music) and video clips (a visual representation of the synthesis of a particular chemical, how to tag a willy-wagtail with a microchip, a snippet from the movie you are analysing (with permission, of course)).

Back to basics

The thesis begins with an excellent introductory chapter that reviews the background literature in an intelligent, scholarly and accurate manner. The candidate takes a commendable 'back-to-basics' approach, reviewing some of the very early (and indeed seminal) papers in the field that are now often (regrettably) overlooked.

It's certainly true that examiners scrutinize your bibliography carefully. It's also true that they particularly notice if the bibliography does not include recent work. Often examiners will comment that there is nothing cited past a certain date - say 2002. They rarely comment if there is nothing before a certain date, but they do notice when original work is cited. Going back to original sources is a sign of high quality scholarship, and it can often lead to new interpretations of that original work which were missed at the time it was being read by other scholars.

The earliest reference in my thesis was to a book published in 1932. What's your earliest reference?

The Thesis

I have recently been examining a PhD thesis of a student from another university (in another country). Overall, it was, in my opinion, an adequate dissertation: the candidate had clearly done a lot of work, there were some original contributions (although they were difficult to find in the presentation), and the written style was passable. While this might not sound very positive, my overall recommendation was "Pass, subject to the questions in my report being addressed at the oral examination, and the revisions in my report being undertaken in the thesis." This recommendation was one of the choices I had, and common to all examination procedures I have ever seen, it required ticking a box.

What I found most difficult in this examination process was identifying what the candidate's actual thesis was; that is, the argument that was being developed over 150 or so pages so as to convince me that this new approach to XXX either had promise as a practical system, or that it gave some insight into existing biological systems. The Graduate Research School webpages give the following advice about the thesis (adapted from Anderson, J. and Poole, M. (1994) Thesis and assignment writing John Wiley and Sons, Brisbane):

"A thesis is not the same as a topic to be investigated. While a topic may be the structure of a particular crystal or the novels of Miles Franklin, a thesis is a statement that says, for instance, something about crystals or an author’s novels. The distinction is important since in many ways the statement of a thesis determines the approach and stance writers take towards their topics or research questions.

A thesis is an idea or theory that is expressed as a statement, a contention for which evidence is gathered and discussed logically. The statement nearly always begins with the word that.

…that the move towards republicanism in Australia began at the end of the last century.

…that use of sapphire as a test mass material in a gravitational wave detector will significantly reduce thermal noise.

…that failure of apparent pregnancies in sheep in the Merredin area is due to high temperatures at the time of mating.

A thesis is a sustained argument. The thesis statement usually appears in the first chapter where the background to the study is described. It will often determine the framework for the literature review and the data collection and it is revisited in the last chapter. It permeates the whole study."

Since I have challenged the candidate I am examining to state his thesis, I now offer the same challenge here: can you state your thesis, in one phrase, starting with the word "that"? 

Theses, books, and articles

Frequently examiners comment on the candidate's publications, or the potential for publications. Here is the gist of three independent comments on the same piece of work; these comments provide an interesting calibration for measuring the thesis against the book or against a set of articles.

In terms of publication, this author should move relatively quickly to approach publishers. The topic is one that lots of people are working on and the data require rapid publication to remain relevant - if not as a complete book, then as a series of articles.

One can typically only write one book based on a PhD thesis.

There is no doubt that there are three or four substantial journal papers that can be mined from this thesis.

The oral exam

Some examiners assume that there will be a viva.

Overall, the experimental methodologies and statistical analyses applied are appropriate for the research undertaken, and the thesis is easy to read and understand. However, the candidate has used different symbols to denote the same variables in different parts of the thesis. The candidate needs to explain the differences between these symbols during the oral examination.

During the oral exam the candidate should report on the status of the papers submitted to journals.

Self plagiarism and salami slicing

Self plagiarism occurs when you re-use your own published work in your current work, without citing the original published work. It's considered an academic crime because in the act of publication you have assigned the ownership of your original work to someone else (the Journal), and now you are passing it off as it it were once again your own.

Salami slicing refers to the act of chopping your work up into tiny publons: units of publication that are minimal at best, just enough to get published (a publon is also known as a least publishable increment). Salami slicing usually occurs with the re-use of data, or with a slight change of data but saying essentially the same thing in another publication. 

Both self plagiarism and salami slicing are considered poor practice because they flood the academic community with near identical papers, making it more difficult for scholars to find relevant information.

The very obvious use of cut-and-paste (identical wording for very large parts of the chapter) is disappointing. Perhaps this could have been avoided by structuring the results differently?

The most common mistake

I am sure it must be true: the most common mistake in English writing is to use "it's" as a possessive.  I know we are all taught this in primary school. "It's" means "it is". Nevertheless, it is a common comment from examiners to PhD candidates:

 "You have made the most common mistake in English writing here. "It's" is a contraction of "it is" and that is the only time an apostrophe is used. "It's a beautiful day" is correct, whereas "The car lost it's control on the slippery pavement" is incorrect. The way to tell whether to use "it's" or not is to substitute "it is" and see if it makes sense. Remove the apostrophe here and wherever else it may be used incorrectly."

Now I know we all know this; nevertheless, this is one of the most common comments made by examiners. Perhaps the linguists among us will tell us whether this confusion actually is the most common mistake in written English. When I checked the web I found a wide variety of claims as to what is the "most common mistake in written English": Your vs. You're; It's vs. Its; There vs. Their; Affect vs. Effect; and the Dangling Participle were high on the list. Also popular were missing articles and spelling mistakes.

The usual convention in scholarly writing is to avoid all use of contractions. The use of contractions creates a familiar tone and does not sound as serious as spelling out the words in full.
 

Proofreading

I read an examiner's report today with 22 pages of minor corrections! It's worth proofreading carefully.

When "not perfect" is an asset

We sometimes worry when our research doesn't turn out the way we wanted it to, or that the results don't seem clean or easy to explain. One examiner didn't find this a problem at all.

 "I particularly like the inherent honesty of her writing, highlighting various difficulties throughout, including her vision for future work and recommendations."