srh.kway.full: Full pipeline: rank-based k-way ANOVA + descriptives + post...

In factorH: Multifactor Nonparametric Rank-Based ANOVA with Post Hoc Tests

Full pipeline: rank-based k-way ANOVA + descriptives + post hocs

Description

Runs a complete nonparametric, rank-based workflow for factorial designs: (1) SRH-style ANOVA table, (2) compact descriptive stats with global ranks, (3) Dunn-Bonferroni post hoc matrices for all effects, and (4) simple-effects post hocs (Bonferroni within each by-table).

Usage

srh.kway.full(formula, data, max_levels = 30)

Arguments

formula	A formula y ~ A (+ B + ...).
data	A data.frame with variables present in formula.
max_levels	Safety cap for number of levels per factor (default 30).

Details

Choice of the ANOVA engine:

• 1 factor: srh.kway() (KW-like),

• 2 factors: srh.effsize() (SRH 2-way + effect sizes),

• 3+ factors: srh.kway() (general k-way on ranks).

Value

A list with elements:

• anova – ANOVA-like table,

• summary – descriptive stats data.frame,

• posthoc_cells – list of p.adj matrices for all effects (from srh.posthocs), or a string when failed,

• posthoc_simple – list of simple-effect tables (from srh.simple.posthocs); for 1 factor: "[not applicable]",

• meta – list with call, n, factor levels, and empty-cell info (if 2+ factors).

Components that cannot be computed for the given design are returned as the string "[not applicable]"; failures are reported as "[failed] <message>".

Examples

data(mimicry, package = "factorH")
# 1 factor
f1 <- srh.kway.full(liking ~ condition, data = mimicry)
# 2 factors
f2 <- srh.kway.full(liking ~ gender + condition, data = mimicry)
# 3 factors
f3 <- srh.kway.full(liking ~ gender + condition + age_cat, data = mimicry)

factorH documentation built on Sept. 11, 2025, 9:09 a.m.