One hypothesis at a time

Pick one hypothesis and name its outcome (what you measured) and its predictor (what you manipulated or compared). Run through this once per hypothesis.

    At a glance

    OutcomePredictor and designTestEffect sizeIf assumptions fail
    Continuous2 groups, different peopleWelch’s td [CI]Mann–Whitney
    Continuous2 conditions, same peoplePaired tmean difference [CI], dzWilcoxon signed-rank
    Continuous3+ groups, different peopleOne-way ANOVA, then Tukey or planned contrastsη²Kruskal–Wallis
    Continuous2+ factors, different peopleTwo-way (factorial) ANOVApartial η²Regression with robust SEs
    ContinuousSame people in several conditions, or nested dataMixed-effects modelβ [CI]Friedman (one factor)
    ContinuousA continuous predictor, or several predictorsLinear regressionβ [CI], R²Spearman (one predictor)
    Two categoriesGroupsχ² (Fisher’s exact for small counts)φFisher’s exact
    Two categoriesContinuous or several predictorsLogistic regressionodds ratio [CI]—
    3+ categoriesGroupsχ² on the full tableCramér’s VFisher’s exact
    One ordinal itemGroupsMann–Whitney (2) or Kruskal–Wallis (3+)rank-biserial r—

    Notes