Measurement Equivalence/Invariance of Scales Measuring Cultural Value: A Proposed Paradigm of Etic-Item Comparison

Abstract

Scholars have recognized that instruments measuring cultural value may not achieve a traditionally accepted level of reliabilities and overall model fit when subjected to confirmatory factor analysis. The poor fit of the configural models intesting measurement equivalence/invariance makes multi-group comparison impossible to continue in the very earlystage. The present solutions provide little utilities in dealing with this issue. In this paper, we propose a paradigm of etic-item comparison with two approaches, namely the prioritizing group-fit approach and the maximizing etic-itemapproach, to solve this problem. We also demonstrate the procedures and outcomes of the etic-item comparison with anempirical example. In the discussion, we address the relevant methodological issues to conclude that the procedures andtools proposed are useful in both the development of scales and the empirical testing of established scales