TheMethods Bench
Start here

Analyse and interpret results

Analysis problems are usually interpretation problems. Start from what you are estimating, choose an approach that fits your variables and design, then report the estimate, its uncertainty and what it means in practice.

Start with this

Choose an analytical approach

Start from your objective and outcome, and get a likely analysis, the effect estimate to report and the assumptions to check.

Open it

Then work through these, in order

  1. Choose the right statistical testWork from variables and design rather than memorising a table.
  2. Know what your variables are doingIndependent, dependent, confounding and mediating variables in practice.
  3. Read p < 0.05 correctlyWhat a small p-value does and does not tell you.
  4. Interpret a 95% confidence intervalReport precision, not just significance.
  5. Judge whether a result mattersSeparate statistical significance from practical importance.
  6. Report results without a p-value listWrite a results section a reader can follow.
Need more support?

Analysis and interpretation support

Paid one-to-one support to align your analysis, outputs and interpretation with the study question.

Scope analysis support