Skip to contents

Choosing a coefficient

An interactive decision helper that recommends which ICC to report and emits the exact icc() call that computes it — no fitting, teaching-first.

choose_icc() format(<icc_recommendation>) print(<icc_recommendation>)
Recommend an ICC and the call that computes it

Estimating an ICC

The workhorse. Fit variance components with a modern mixed model and compute the intraclass correlation for a stated design: absolute agreement vs. consistency, single vs. average, fixed vs. random raters, one-way vs. two-way, multilevel (subject vs. cluster level, crossed or nested), and incomplete (ragged) data — all with boundary-aware Monte-Carlo confidence intervals. The print(), summary(), tidy(), glance(), autoplot() (coefficient forest / variance-component decomposition) and plot() methods are documented on this page.

autoplot.icc() plot(<icc>) format(<icc>) print(<icc>) summary(<icc>) tidy(<icc>) glance(<icc>) icc()
Intraclass correlation coefficient for a two-way design

Decision (D-) studies

Project a fitted ICC’s reliability to other numbers of raters and plot the curve, plus its tidy()/glance()/autoplot()/plot() methods.

Tidy methods

Broom generics re-exported so tidy() and glance() resolve without attaching the generics package.

reexports tidy glance
Objects exported from other packages

Datasets

Worked rater-reliability examples used across the docs and vignettes.

ratings
Rater reliability example (Shrout & Fleiss, 1979)
ratings_incomplete
Rater reliability example with missing cells

Package overview

intraclass intraclass-package
intraclass: Modern Intraclass Correlation Coefficients