Chop and Change: A Commentary and Demonstration of Classical vs. Modern Measurement Models for Interpreting Latent-Stability of Occupational-Future Time Perspective

This commentary article was initially motivated by an empirical paper published in the journal of Work, Aging, and Retirement that reported support for stability (non-decreasing) future time perspectives (FTP) over two repeated-measurements. That is, empirical evidence supporting the temporal stability of an adapted measure (occupational-FTP [O-FTP]) serves as guiding framework for demonstrating limitations of classical test theory (CTT) and modern psychometrics’ (IRT) enabling extension for stronger substantive inferences from response data. The focal authors’ quantitative attention to study design and statistical analysis is commendable. In this commentary, I aim to complement their efforts from a measurement perspective. This is accomplished through four sections. In the first section, I summarize some well-known limitations to CTT measurement models for assessing change. Then, I briefly introduce item response theory (IRT) as an alternative test theory. In the second section, Chop, I review the empirical evidence for FTP and O-FTP’s latent-factor structure. Then, I bring evidence from modern psychometric methods to bear on O-FTP, specifically, a model-comparisons approach was adopted for comparing relative fit of 1-factor, 2-factor, and bifactor solutions in cross-sectional data (N = 511). Findings supported retention of the bifactor solution. In the third section, Change, I extend the bifactor model to two-wave FTP data over approximately 2 years (N = 620) as an instructive application for assessing temporal stability. The fourth section concludes with a brief discussion of substantive implications and meaningful interpretation of (O)-FTP scores over time.