pg-mean-testslisted

# PG Mean Tests Use for t tests and post hoc pairwise comparisons where the dependent variable is approximately continuous. ## Load Read: - `../../references/supervision-gates.md` - `../../references/pingouin-api-quickref.md` - `../../references/pingouin-optimization.md` - `../../references/apa-output-template.md` if writing results. ## Decision Rules - One sample against a known value -> `pg.ttest(x, y=<value>)`. - Two independent groups -> `pg.ttest(x, y, paired=False, correction="auto")`. - Same participants measured twice -> `pg.ttest(x, y, paired=True)`. - More than two group levels or multiple pairwise contrasts -> `pg.pairwise_tests`. - Non-parametric pairwise comparisons -> `pg.pairwise_tests(..., parametric=False)`. ## Required Inputs - Outcome column. - Group or condition column. - Subject ID for paired/repeated comparisons. - Which comparisons are planned versus post hoc. - Multiple-comparison correction: default to `holm` for post hoc families unless the user specifies another correction. - Alternative hypothesis: default to `two-sided`. ## Code Patterns Independent t test: ```python x = df.loc[df["group"].eq("A"), "score"] y = df.loc[df["group"].eq("B"), "score"] res = pg.ttest(x, y, paired=False, correction="auto", alternative="two-sided", confidence=0.95).round(3) pg.print_table(res) ``` Paired t test from wide columns: ```python res = pg.ttest(df["pre"], df["post"], paired=True, alternative="two-sided", confidence=0.9