There is a format for the thesis and it goes like this – Introduction, Literature review, Methods, Report of research findings, Discussion and conclusion. This is known as ILMRaD. Many theses follow this structure and it does ‘work’ – it’s the default textual position, the thesis equivalent of Times New Roman.

But it’s not the only way to construct the thesis text.

Now I’m not suggesting here that you can abandon all of the content that ILMRaD encompasses. You can’t. A thesis does need to provide a justification for the study, locate it within a field of literature and maybe policy and/or practice. It does need to say how the research was designed and conducted, and provide an audit trail for the examiner. There clearly needs to be writing about the research findings themselves. The conclusion does need to say not only what the thesis was about and did, but also argue this as a contribution to knowledge and maybe policy and/or practice.

But having to include all these things isn’t the same as using them as the structure of the text.

Doctoral researchers often ask me if they need a literature chapter per se. Well the answer is not necessarily. If it’s OK with the supervisor, and it suits the argument that is being made, the literature discussion may well be divided and spread around the text. In my own thesis I had to provide a policy context, and a history of the field of research, and a history of practice – and all of this used literatures very extensively, and it made up the first three chapters. I just couldn’t have divided the literatures up in any sensible way and made the argument I needed to if I had used a single chapter (1).

Doctoral researchers also often assume that they must provide a single methods chapter. Yet several doctoral researchers I work with have conducted two phases of study, with a policy analysis leading to case study. It actually made more sense for (and to) them to first of all discuss the method of policy analysis and show the results, because that led directly on to a set of decisions about case studies. The methods used for the case studies and the cases were then presented as a phase two. It was helpful for the examiners and other readers to follow the two phases one after the other, to see the logic of the research represented in and as the text.

My view is that it’s important to follow the logic of the process and the argument and use that to sort out how the text goes, and THEN write the mandatory stuff as it works for the argument and the reader. This does of course require some deep and sometimes difficult thinking about the text and the argument – and some creativity. This approach is not about slotting sections into a pre-prepared structure. So deviating from the default might be seen as harder work than using it – and more risky. However, doctoral researchers whose work doesn’t neatly fit into ILMRaD often find it much harder to try to meet the default than to invent a structure that works for their study.

But here’s a warning. Some examiners are VERY hostile to varying the ILMRaD format.

One of the doctoral researchers I’ve worked with had one of those rather difficult examination experiences. She’d done her research and the usual kind of analysis, and didn’t think much of what she had. She then alighted on a theoretical approach which required her to do something different with the data – a second analysis. She wanted to tell the story of her learning through the thesis and structured the text into the two stages she’d experienced. However one of the examiners would not accept this and required a retro re-write which made the thesis represent the finished picture, rather than the journey that the doctoral researcher had wanted. Both the researcher and I were put out, I have to say. We both thought that presenting a text as a representation of the actual learning was OK. But clearly not all examiners agree.

The moral of this story is that if the thesis has moved away from ILMRaD it is absolutely vital to check out, when selecting them, how the examiners feel about not using the default. (In the North American committee system, permission to move away from ILMRaD is generally part of the ongoing negotiation of the research, rather than something more chancy at the end of the process.)

And some disciplines and some countries seem to be more wedded to ILMRaD than others. My sense of it is that Australian theses in the humanities and social sciences genuflect less to the traditions of ILMRaD than is the case in the UK – but I’m only talking from my experience here of course, and it would take a more systematic study to verify whether this was the case. I also get the sense that in some disciplines there is more willingness to experiment with genres other than the Big Book. And maybe the thesis as monograph is under duress anyway from shifts towards the doctorate by publication and portfolio.But in the meantime, I think it is worth holding the default ILMRaD up to some scrutiny.

Like all defaults, ILMRaD does work. But it IS a default, and it isn’t the only way to construct a thesis. The format will certainly get the thesis written. However, it won’t stand out from the crowd – and it may not actually do the research justice.

Note
(1) Did I suffer for not doing ILMRaD? No. The examiners were happy and there were no corrections. The thesis also won the annual national award in my field – so scholarly community approval for doing something internally logical and coherent. But my thesis was written in the days before word limits – and I did use a lot of words. So not a good current exemplar!