Linking Error in the 2PL Model

The two-parameter logistic (2PL) item response model is likely the most frequently applied for analyzing dichotomous data. Linking errors quantify variability in means or standard deviations due to choice of items. Previous research presented analytical work linking one-parameter model. In this article, we present 2PL using general theory M-estimation. are derived case log-mean-mean two groups. performance newly proposed formulas evaluated a simulation study. Furthermore, error estimation also treated more complex settings, such as chain linking, trend estimation, fixed parameter calibration, and concurrent calibration.