Article contents
In Praise of Pluralism. A Comment on Borsboom
Published online by Cambridge University Press: 01 January 2025
Extract
I tend to agree with Professor Borsboom that psychology, and more generally the social sciences, could benefit from better psychometric modeling. However, if psychometric developments are to have more effect on everyday practice in psychology, psychometricians probably need to pay more attention to the substantive and methodological problems in various areas of psychology. For example, Professor Borsboom is critical of the Standards for educational and psychological testing (AERA, APA, NCME, 1999) for suggesting that group differences in testcriterion relationships are relevant to test bias. He bases his criticism on the finding that predictive invariance is not the same as measurement invariance. However, he fails to acknowledge the social, political, and ethical problems associated with failures of predictive invariance in high-stakes contexts (e.g., employment and admissions testing).
- Type
- Original Paper
- Information
- Copyright
- Copyright © 2006 The Psychometric Society
References
- 6
- Cited by