Formative Versus Reflective Measurement in Executive Functions: A Critique of Willoughby et al

摘要

Research into executive functioning (EF) has indeed grown exponentially across the past few decades, but as the Willoughby et al. critique makes clear, there remain fundamental questions to be resolved. The crux of their argument is built upon an examination of the confirmatory factor analysis (CFA) approach to understanding executive processes. Beyond the methodological critique, questions about the study of individual differences in EF surface throughout their thoughtful review and analysis. Below we consider each of these topics in turn. The more distant reader may be surprised by the lack of agreement among EF researchers about basic theoretical and methodological issues. There is widespread agreement that EF refers to a suite of processes mediated by the prefrontal cortex that support goal-directed actions (Diamond, 2013), but beyond this general description, disagreement emerges. Perhaps the most difficult challenge for EF research has been the problem of task measurement (e.g., poor intertask correlation, lack of precision in targeting specific EF processes or "task impurity"). The CFA approach, yielding latent traits, has provided one resolution to the problems associated with task impurity. The seminal Miyake et al. (2000) has been hugely influential, giving rise to widespread acceptance of the tripartite model. Again, it is easy to find fundamental disagreements in EF research (see, for example, the disagreement about the distinction between working memory and inhibitory control, reviewed in Diamond, 2013), but it is fair to say that the original Miyake et al. (2000) CFA gave rise to a widely accepted framework. Since then, the CFA approach has been implemented to explore new directions in EF research, such as the emerging "hot" versus "cool" distinction that we believe is central to the exploration of individual differences (Welsh & Peterson, 2014).