Measurement error significantly biases interaction effects and distorts researchers' inferences regarding interactive hypotheses. This article focuses on the single-indicator case and shows how to accurately estimate group slope differences by disattenuating interaction effects with errors-in-variables (EIV) regression. New analytic findings were…

Presents minimization and maximization models for parallel test construction under constraints. The minimization model constructs weakly and strongly parallel tests of minimum length, while the maximization model constructs weakly and strongly parallel tests with maximum test reliability. (Author/SLD)