Evaluating the Invariance of Cognitive Profile Patterns Derived from Profile Analysis via Multidimensional Scaling (PAMS): A Bootstrapping Approach

摘要

The aim of the current study is to validate the invariance of major profile patterns derived from multidimensional scaling (MDS) by bootstrapping. Profile Analysis via Multidimensional Scaling (PAMS) was employed to obtain profiles and bootstrapping was used to construct the sampling distributions of the profile coordinates and the empirical confidence intervals for the coordinates. The PAMS approach uses ordinary MDS, but considers dimensions as the most typical (or major) profiles in a population. Since the PAMS approach is essentially exploratory, its outcomes need to be confirmed: The study analyzed one sample to uncover the major profile patterns and used another sample as a cross-validation study. The empirical confidence intervals of both samples were compared to examine the invariance of the profile pattern structures. The bootstrap results supported the invariance of the pattern structures. The educational applications and implications about the PAMS research are discussed.