Comments on van der Linden's Critique and Proposal for Equating
Journal of Educational Measurement
Published online on September 22, 2013
Abstract
While agreeing with van der Linden (this issue) that test equating needs better theoretical underpinnings, my comments criticize several aspects of his article. His examples are, for the most part, worthless; he does not use well‐established terminology correctly; his view of 100 years of attempts to give a theoretical basis for equating is unreasonably dismissive; he exhibits no understanding of the role of the synthetic population for anchor test equating for the nonequivalent groups with anchor test design; he is obtuse regarding the condition of symmetry, requiring it of the estimand but not of the estimator; and his proposal for a foundational basis for all test equating, the “true equating transformation,” allows a different equating function for every examinee, which is way past what equating actually does or hopes to achieve. Most importantly, he appears to think that criticism of others is more important than improved insight that moves a field forward based on the work of many other theorists whose contributions have improved the practice of equating.