stepwise_pcv: Stepwise Proportional Change in Variance (PCV)

In MAIHDA: Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy

Stepwise Proportional Change in Variance (PCV)

Description

Estimates the proportional change in variance (PCV) sequentially by fitting intermediate (partially-adjusted) models. It adds each predictor variable one-by-one to gauge its unique contribution in explaining between-stratum inequalities.

Usage

stepwise_pcv(data, outcome, vars, engine = "lme4", family = "gaussian")

Arguments

data	Data frame with observations. Ensure 'make_strata()' was run first so the 'stratum' variable exists.
outcome	Character string; the dependent variable.
vars	Character vector; predictors (strata groupings & covariates) to add sequentially to the model.
engine	Modeling engine ("lme4" or "brms"). Default is "lme4".
family	Error distribution and link function. Default is "gaussian".

Value

A data.frame showing the sequential models, the between-stratum variance at each step, and both the step-specific and total PCV.

Examples

strata_result <- make_strata(maihda_sim_data, c("gender", "race"))
stepwise_pcv(strata_result$data, "health_outcome", c("gender", "race", "age"))

MAIHDA documentation built on April 5, 2026, 5:06 p.m.