Skip to content

Our methodology

Good research should be inspectable. A research output is only as strong as the decisions behind it, and the question, design, data, analysis, interpretation and reporting all have to fit together.

How the work is built

Seven decisions, in order.

Each one constrains the next. Taking them out of order is how a project ends up with an analysis that cannot answer the question it was built for.

  1. Question

    We clarify what the study is actually trying to answer. A vague question produces vague methods, weak interpretation and reviewer problems that surface far too late to fix cheaply.

  2. Design

    We match the design to the question, whether the project is an evidence synthesis, an observational study, a statistical analysis or a manuscript already in progress.

  3. Data or evidence

    We examine what information actually exists and whether it can support the intended conclusion. This is the step where a study is most often rescued, by being redefined before it is run.

  4. Analysis

    Methods follow from the question, design, outcomes, assumptions and data structure. The statistical test is the last decision, not the first.

  5. Interpretation

    We say what the results mean, what they do not mean, and which limitations have to be stated. A confident sentence does not make weak evidence strong.

  6. Reporting

    Reporting is aligned to the appropriate standard so another researcher can follow the decisions that were made, rather than having to trust that they were reasonable.

  7. Reproducibility

    Where applicable we deliver documented R or Stata workflows, so the analysis can be inspected, rerun, extended and defended without rebuilding it from scratch.

Why physician-led

The methodology has to understand the medicine.

Choosing an outcome, a comparator, a population or an analysis is a clinical decision as much as a statistical one. A statistician can pick the wrong test for the right question; a clinician can pick the right outcome and the wrong method. The work sits at that intersection deliberately.

Tell us about your study

Send the scope and we will come back with a written quote. No account needed.