Use this guided checker to examine the structure of a planned or completed study, identify common design-analysis mismatches, and see which assumptions or dependencies need attention. This beta currently provides full methodological guidance for continuous numerical and binary outcomes.

Illustration of a study cohort divided into comparison groups, followed across repeated measurements, and summarised with effect estimates.

Before you begin

What this checker provides

Answer a short sequence of adaptive questions and receive a structured, rule-based report tailored to the study design you describe. The checker is designed to flag important methodological issues, suggest potentially suitable approaches within its supported scope, and make clear when more information or specialist review is needed.

Your report will include

  • A study-structure summary covering the outcome-bearing or experimental unit, group structure, dependence, repeated measurements, clustering, missingness, and multiplicity.
  • Potentially suitable analytical approaches for studies within the checker’s supported scope, with a concise explanation of why they may fit.
  • Warnings and assumptions that should be checked before results are relied upon.
  • A reporting checklist and an explanation of any additional information needed for a stronger statistical assessment.

How it can help

  • Spot design–analysis mismatches, overlooked dependencies, replication problems, or multiplicity concerns before analysis begins.
  • Review whether a reported method broadly reflects the study structure and identify points that need verification.
  • Prepare a clearer analysis plan or Methods section and a focused set of questions for a statistician.
  • Identify when the available information or study structure requires case-specific review.

The questions adapt to your answers, and a typical pathway is designed to take only a few minutes. The resulting report can be printed or downloaded as a PDF.

The checker does not analyse raw data, verify that an analysis was implemented correctly, or replace case-specific advice from a trained statistician. For some studies, “specialist review advised” or “insufficient information” is therefore the appropriate result.