Intraclass correlation (ICC) in multilevel models

This fact sheet explains the intraclass correlation coefficient (ICC) as a measure of within-unit dependence in repeated-measures and multilevel data. It describes how the ICC is derived from variance components in random-intercept linear mixed models, how it should be interpreted in applied research, and why it directly affects statistical precision, power, and valid inference. The document provides practical guidance for researchers on recognising non-independence, choosing appropriate analytic models, and interpreting ICC values in longitudinal and clustered study designs.