A Mediated Measurement Model to Investigate Whether Test Scores Are Invariant to the Test-Taking Strategies

• In Event: Marriott Roundtable Session 41
  In Roundtable Session: Modeling and Detecting Measurement Invariance

Abstract

This paper proposes to relax the constraint that the variable tested for measurement invariance (MI) has to denote a pre-existing demographics, because 1) the constraint is not inherent in the general mathematical definition of MI and 2) doing so promotes flexible application of MI, which is the focus of the current paper. We discuss how a non-pre-value-judged MI allows for a latent, continuous MI variable that can be incurred by the measurement process. Furthermore, we provide a mediation model as one way to test it and conclude with a working example of understanding test-taking strategies as a proof of concept.

Authors

• Amery D. Wu, University of British Columbia

• Shun-Fu Hu, The University of British Columbia

• Michelle Yue Chen, Paragon Testing Enterprises / The University of British Columbia

• Lok Heng Chau, The University of British Columbia