Home/Insights/Methodology reference
Research methodology reference
27 reference pages on research design. Each states what the design can establish, what it structurally cannot, the threats that most often undo it, and the reporting guideline that governs it.
The limits are the point. Most over-claiming in a manuscript happens in the discussion, where a conclusion outruns the design that produced it — so each page says plainly where that line falls.
Foundations 7
Methodology in researchMethod is what you did. Methodology is why that was the right thing to do. Papers are rejected for confusing the two.The scientific approach in researchHypotheses are tested by trying to refute them, which is why a study designed only to confirm an expectation tells you very little.Types of scientific researchThe useful classification is not basic versus applied. It is what kind of claim the design can support.Primary research methodsPrimary research generates data that did not previously exist, which means the researcher owns every design decision — and every bias introduced by one.Secondary research methodsUsing data someone else collected is fast and cheap, and the constraint is permanent: you cannot add the variable that was not recorded.Desk study methodologyA desk study answers a question from existing sources. Its credibility rests entirely on how the sources were found and chosen.Evaluating scientific researchCritical appraisal asks whether a study’s results are believable, which is a different question from whether they are interesting.
Observational designs 6
Cohort study methodologyA cohort follows people forward from exposure to outcome. It is the observational design that gets closest to a causal claim, and the one most often undone by who left.Cross-sectional research designEverything measured at once. Excellent for prevalence, structurally unable to tell you what came first.Longitudinal research designRepeated measurement on the same people. It can separate change within a person from differences between people — provided the people stay.Correlational research designIt measures whether two variables move together. That is a genuine finding and it is not a causal one, however strong the coefficient.Descriptive research designIt characterises what is there. Without a comparison group it cannot support any statement about why, and the discussion is where that gets forgotten.Comparative study methodologyTwo or more groups compared without random assignment. Everything depends on how unlike the groups were to begin with.
Experimental and quasi-experimental 5
Experimental design in research methodologyThe investigator controls the exposure and assigns it at random. That one feature is what licenses a causal claim.Quasi-experimental methodologyAn intervention without random assignment. Causal inference is possible, but only on assumptions that have to be stated and defended.Quasi-experimental design: advantagesWhere randomisation is impossible, the realistic comparison is not with a trial — it is with having no evidence at all.Quasi-experimental design: disadvantagesEvery limitation reduces to one thing: the groups were not formed at random, so any difference between them may be the reason for the result.Response surface methodologyA designed-experiment approach for finding the settings that optimise a process, using far fewer runs than changing one factor at a time.
Qualitative and mixed methods 6
Case study research methodologyAn in-depth investigation of one case in its real context. Its generalisation is to theory, not to a population, and conflating the two is the usual criticism.Qualitative case study methodologyWhere the case study tradition meets interpretive enquiry: the aim is understanding the case from the inside, and the methodology has to say whose understanding.Reflexivity in qualitative researchThe researcher shapes the data they generate. Reflexivity is the discipline of examining that influence, and reporting it — not a paragraph apologising for bias.Mixed methods researchQuantitative and qualitative strands in one study. It is only mixed methods if the strands are integrated; running both and reporting them separately is two studies in one paper.Psychology research methodsPsychology’s methods are shared with much of the life sciences; what is distinctive is how publicly the field has had to confront what happens when they are applied loosely.Empathic design researchObservation-led design research that looks for the needs people cannot articulate — and that must be disciplined, or it becomes the designer’s intuition with fieldwork attached.
Applied and consensus methods 3
Delphi study methodologyStructured, anonymous, iterative expert consensus. The design stands or falls on decisions that are too often left unreported: who the panel were, what consensus meant, and when to stop.Feasibility study methodologyIt asks whether the definitive study can be run, not whether the intervention works — and reporting an effect estimate as though it answered the second question is the standard error.Impact study methodologyImpact evaluation asks what would have happened otherwise. Without a credible counterfactual, it measures change and calls it effect.