Coherence in research design is not the same as coherence in academic writing, although the two are related.

Research design coherence is sometimes described through the metaphor of a red thread, one continuous line of logic running from the question through every subsequent decision. This metaphor is a just a little bit too tidy, research is messier than a single thread suggests. But the idea of a something threading through a research project, stitching it down and holding it together, is helpful.

Every significant decision in a research design hinges on how the researcher understands what they are doing. How knowledge is produced. What counts as a legitimate question. What kinds of evidence can answer it. What relationship the researcher has to the people and phenomena being studied. What analytical processes will work on the material generated. These decisions need to be consistent with each other. When they are not, the research design is incoherent, and no amount of elegant writing about individual components will fix it.

One common research design coherence problem is a mismatch between the stated methodology and the actual methods. A researcher says they take a phenomenological approach – they are concerned with the lived experience of participants, with bracketing researcher preconceptions, with describing rather than explaining. And then they design an interview schedule full of closed questions, code the data into predetermined categories, and write results that conform to a theory. The methodology says one thing; the methods another.

A related problem is when a theoretical framework floats free of the analysis. The writer spends a chapter establishing a theoretical position and then conducts an analysis that does not use it. The theory appears in the literature review and again in the conclusion, bracketing an empirical middle section that could have been written by someone with a completely different theoretical orientation. The theoretical framework was decoration rather than underpinning the study. The design looked coherent on paper; in practice the theory and empirical components were running in parallel rather than together.

Ontological and epistemological positioning generates a version of incoherence too. Most doctoral researchers are expected to explain their ontological and epistemological assumptions. The problem arises when explanations are written without looking at the implications for everything that follows. A researcher who identifies as a social constructionist but then treats their interview data as a transparent window onto participants’ inner lives has a coherence problem that starts at the foundations. The declared position and the actual practice are inconsistent, and the inconsistency ripples outward through every subsequent methodological decision.

Researchers cannot  easily retrofit coherence in research design. It’s harder than making writing coherent. With writing you can revise: clarify the argument, cut the sections that don’t serve it, rewrite the transitions. But research design coherence problems, identified late, can require rethinking the study at a level that feels catastrophic at year two or three of a project. This is one reason why doctoral supervisors push you really hard on research design questions early, and why it is worth treating that pushing seriously rather than see it as an obstacle to you getting on with the interesting work, the ‘real’ research. That interesting work absolutely depends on the research design being sound and defensible.

One strategy to ensure research design coherence is to hold the whole design in view at once and ask, repeatedly, whether the components are making consistent claims. One way to do this is to write a short account, a page, maybe two, about the logic of your study. What you are asking, why that question requires a particular kind of knowledge, what methodology follows from that, what methods follow from that methodology, and what you will be able to claim as a result. Not a chapter-by-chapter summary of the thesis, but a compressed account of the reasoning. If you can write that account and the logic holds throughout, the design is probably coherent. If you find yourself unable to connect one stage to the next without a leap of faith, then there is more work to do.

Coherence in research design is about the researcher being clear with themselves, then with their readers, about what kind of knowledge their study produces, and making every methodological decision consistent with that. A coherent design does not claim more than its methods can deliver, does not use theory as decoration, and does not declare an epistemological position and then quietly abandon it when it becomes inconvenient. Research coherence is what makes a study trustworthy, and trustworthiness is what makes findings worth attending to.

Photo by Andrea Seiler on Unsplash