Designing a PhD: Study Design and Methodology

The design of a PhD study is crucial to its success. The design process varies enormously according to the discipline. In the physical sciences, you are governed by the principles of scientific method. Study design may focus on detailed decisions about how you design your experiments and apparatus. In some cases, the contribution to knowledge may actually be in building an apparatus which enables a new experiment to be done.

In medical studies, the study design often involves selecting from a range of prescribed trial methodologies. Broadly categorised as randomised control trials, these methods seek to remove confounding factors by constructing an active group and a control group where the confounding factors would be randomly distributed across both groups. In reality, the student will need to identify the particular flavour of trial that meets the needs of their study.

In the social sciences and humanities, it is much less likely that a prescribed study design will exist and the choice and justification of your study design is likely to be a very important part of your study. I have met physical science PhDs from very respectable Universities who when asked “Why did you do your study that way?” would answer “Because my supervisor told me to!”. This is not likely to go down well in a social science context, where very often in my experience, discussion of methodology at examination time lasts longer than discussion of the results themselves.

From an examiner's perspective, it provides very good evidence of the level of understanding of the student, and givesconfidence in the results.

In reality, the study design will almost certainly be based upon an existing approach or approaches. Therefore, the first step is to identify the possible study design approaches. At the most basic level, are you seeking to generate numerical or qualitative data, or perhaps a combination of the two? Once possible methods have been identified, then you can either adopt an existing approach, or adapt it for your own study. You may also need to combine different methods within your overall design, especially if you are intending to use a mixture of qualitative and quantitative approaches.

A PhD is quite a big piece of work. Therefore your study design may well need to be constructed in phases. For example, at my University, a student will enrol on an MPhil programme with a view to transferring to the PhD once the student has demonstrated that they have the skills needed and that their project is viable at PhD level. This creates a natural structure for the study design. The first part (leading to MPhil or transfer to PhD) is generally concerned with analysing an existing problem situation, gathering prior knowledge and baseline data on the problem. It allows the student to demonstrate their ability to carry out independent research. This phase will often conclude with proposing a solution to address the problem. The PhD phase is then concerned with developing and evaluating the proposed solution.

I deliberately described a PhD as quite a big piece of work. That’s because students often overestimate what can be achieved. Once you start to study a problem, it can start to grow in scope quite alarmingly, so it is important to start focused. Another problem is that some classical research approaches are simply too big for an individual PhD, so unless your PhD is part of a team project you may not be able to use them. For example, many medical trials last longer than a typical PhD.

On the qualitative side, the grounded theory approach is popular with qualitative researchers. However, it is in essence an iterative approach, and many PhD studies do not allow enough iterations to truly be regarded as implementations of the classical grounded theory approach. On the other hand, it is perfectly valid to use a design based upon grounded theory and then discuss in the reflections and conclusions at the end whether the approach taken truly captures all the features of grounded theory as a limitation of the study.

At the end of your PhD, you should always reflect on how well your chosen design served your needs. A good starting point for these reflections is to consider which bits of the study design would you do differently if you had your time again and why.