Discrete-state models: Comment on Pazzaglia, Dube, and Rotello (2013).

Abstract

Pazzaglia, Dube, and Rotello (2013) have provided a lengthy critique of threshold and continuous models of recognition memory. Although the early pages of their article focus mostly on the problems they see with 3 vintage threshold models compared with models from signal detection theory (SDT), it becomes clear rather quickly that Pazzaglia et al. are concerned more generally with problems they see with multinomial processing tree (MPT) models. First, we focus on Pazzaglia et al.'s discussion of the evidence concerning receiver operating characteristics (ROCs) in simple recognition memory, then we consider problems they raise with a subclass of MPT models for more complex recognition memory paradigms, and finally we discuss the difference between scientific models and measurement models in the context of MPT and SDT models in general. We argue that Pazzaglia et al. have not adequately considered the evidence relevant to the viability of the simple threshold models and that they have not adequately represented the issues concerning validating a cognitive measurement model. We further argue that selective influence studies and model flexibility studies are as important as studies showing that a model can fit behavioral data. In particular, we note that despite over a half century of effort, no generally accepted scientific theory of recognition memory has emerged and that it is unlikely to ever emerge with studies using standard behavioral measures. Instead, we assert that useful measurement models of both the SDT and the MPT type have been and should continue to be developed.

The two recollections.

“…However, such intermodel correlations may not be strong because the models differ dramatically in the degree to which the processes that they measure are treated as discrete or continuous variables. Here, it has often been hypothesized that models that differ in this way may produce different findings (Pazzaglia et al, 2013), although we saw earlier that intermodel correlations for the six parameters of the unconstrained models were very high (>.90) and that some similar results have been obtained in simulations of simple recog nition models (Batchelder & Alexander, 2013).…”

“…Although the present model is a threshold model, it does not follow that conclusions about old/new recognition apply to other paradigms, such as conjoint recognition, process dissociation, or source monitoring (Batchelder & Alexander, 2013). Nevertheless, we evaluated this possibility by conducting simulations of the model's parameter space under variations in its bias parameters.…”

“…With respect to commonalities, there are three principal ones. The first and most fundamental one is an exten sion of some results that were recently reported by Batchelder and Alexander (2013) and concerns correlations among the models' re spective estimates of the retrieval and bias/criterion parameters. Using Note.…”

The flexibility of models of recognition memory: The case of confidence ratings

“…For the models with response mapping, it is possible, for example, that lowest confidence ratings would be strongly preferred even in detect/recollection states (i.e., these confidence levels can have the largest s parameters). This has prompted criticisms to the effect that the models with response mapping deal with rating scales in an arbitrary and posthoc manner (e.g., ; see also Batchelder & Alexander, 2013). According to the critics, these model variants are overly complex (Dube, Rotello, & Heit, 2011).…”

Characterizing Belief Bias in Syllogistic Reasoning: A Hierarchical-Bayesian Meta-Analysis of ROC Data

“…There is a decade-long debate among SDT and discretestate modelers on the relative merits of the two approaches in several psychological domains (Batchelder et al, 1994;Dube et al, 2010Kellen, 2014;Kellen et al, 2013Kellen et al, , 2015Kinchla, 1994); for reviews, see Pazzaglia et al (2013), Batchelder and Alexander (2013), and Dube et al (2013). 13 From this heated debate, two constructive points are often overlooked: First, there is some consensus that the two modeling approaches seem to be particularly successful in certain types of domains and paradigms.…”