Between Subjects Variance and Within Subjects Variance in ICC

[tkaczye]

Hello Sudhi, indeed this is a fundamental concern with methods like ICC – a measurement in a heterogeneous population appears more reliable than the exact same measurement applied to a homogeneous population. I have essentially stopped using ICCs for this reason, and followed prof Harrell’s advice to rely on U-statistics. Related post:

Mixed vs Random Effects ICC

Thank you for this interesting discussion. We will also look to U statistics in our future papers, but previously have always relied on ICCs as the norm for measuring repeatability. I am often confused by the varying formulas and am glad for the above discussion and helpful article referenced above by Koo and Li to interpret McGraw&Wong’s structure better. Other authors have advocated that a minimal detectable change (MDC95) corresponds to the real world better. This is a bit odd since the MD…