Explain a model's factors and dummy coding in one call

explain_factors() is the recommended entry point for understanding the categorical predictors in a fit. For every factor it states, in plain language, the coding scheme (its levels, which level is the reference, and how many indicator columns it contributes), then shows the per-group means (group_means()) and the pairwise comparisons (group_contrasts()). For every interaction it gives the difference-of-differences reading together with the inline cell means or per-group slopes.

Pass a fitted model (it is run through symbolize() for you) or an existing symbolized_model.

Print the result at the console for a walkthrough, or knit it inside a Quarto / R Markdown document for a heading-and-table section per factor.

Usage

explain_factors(x, ...)

Arguments

x

A fitted model supported by symbolize(), or a symbolized_model.

...

Passed to symbolize() when x is a fit.

Value

A symbolized_factor_explanation object.

See also

group_means(), group_contrasts(), explain()

Examples

d <- transform(mtcars, gear = factor(gear))
if (requireNamespace("emmeans", quietly = TRUE)) {
  # one call: coding scheme, group means, and pairwise contrasts per factor
  explain_factors(lm(mpg ~ gear, data = d))
}
#> 
#> ── Factors in your <lm> model ──────────────────────────────────────────────────
#> 
#> ── gear ──
#> 
#> `gear` has 3 levels (3, 4, 5). R uses 3 as the baseline and adds 2 indicator
#> column(s); each coefficient is the difference from 3, not that group's own
#> mean.
#> 
#> 
#> ── Group means ──
#> 
#> gear=3 estimate = "16.1" (13.6, 18.6) *
#> gear=4 estimate = "24.5" (21.8, 27.3) *
#> gear=5 estimate = "21.4" (17.1, 25.7) *
#> Scale: response. CI method: wald. Rows marked `*` have a 95% interval that
#> excludes zero.
#> 
#> 
#> ── Group contrasts (pairwise) ──
#> 
#> gear3 - gear4 estimate = "-8.43" (-12.2, -4.70) *
#> gear3 - gear5 estimate = "-5.27" (-10.2, -0.301) *
#> gear4 - gear5 estimate = "3.15" (-1.97, 8.28)
#> Intervals are per-contrast, not family-wise; pass adjust = "tukey" for
#> simultaneous bands. Scale: response. CI method: wald. Adjustment: none. Rows
#> marked `*` have a 95% interval that excludes the null.