Extends state-trace analysis (STA) to any arbitrary number of tasks
•Proposes the first rigorous definitions of single- and multiple-systems models
•Shows that STA provides no information about the number of underlying systems
•Shows that STA can test for double dissociations, but not for single dissociations
AbstractState-trace analysis (STA) is a method for determining the number of underlying parameters or latent variables that are varying across two or more tasks. STA is based on the fact that under very weak conditions, any model in which r parameters are varying across r or more tasks predicts an r-dimensional state-trace plot. Although monotonicity assumptions can sometimes be useful in STA, they are not required. Specifically, there is no need to assume that performance in any task is a monotonic function of whichever parameters are varying. As a result, requiring STA models to assume monotonicity seriously reduces the applicability of STA. Whereas an STA can identify the number of varying parameters, it provides no information about the number of underlying systems. Similarly, STA is ill suited to examining dissociations. It can be used to test for double, but not single dissociations. In particular, a monotonic state-trace plot rules out a double dissociation but provides no information about whether or not the data contain a single dissociation.
KeywordsState-trace analysis
Monotonic state-trace model
Single-versus multiple systems
Dissociations
© 2022 The Author(s). Published by Elsevier Inc.
留言 (0)