Search

van Doorn et al. (2021) outlined various questions that arise when conducting Bayesian model comparison for mixed effects models. Seven response articles offered their own perspective on the preferred setup for mixed model comparison, on the most appropriate specification of prior distributions, and on the desirability of default recommendations. This article presents a round-table discussion that aims to clarify outstanding issues, explore common ground, and outline practical considerations for any researcher wishing to conduct a Bayesian mixed effects model comparison.

Memory theories distinguish between item and associative information, which are engaged by different tasks: item recognition uses item information to decide whether an event occurred in a particular context; associative recognition uses associative information to decide whether two events occurred together. Associative recognition is slower and more difficult than item recognition, suggesting that item and associative information may be represented in different forms and retrieved using different processes. Instead, I show how a dynamic model (Cox & Criss, 2020; Cox & Shiffrin, 2017) accounts for accuracy and response time distributions in both item and associative recognition with the same set of representations and processes. Item and associative information are both represented as vectors of features. Item and associative recognition both depend on comparing traces in memory with probes of memory in which item and associative features gradually accumulate. Associative features are slower to accumulate because they emerge from conjunctions of already-accumulated item features. I apply the model to data from 453 participants who each performed item and associative recognition following identical study conditions (Cox et al., 2018). Comparisons among restricted versions of the model identify a limit on the rate at which item features can accumulate from multiple items as well as a limit on the total number of features that can be used to probe memory at one time. The apparent limit on probe feature capacity is interpreted as an adaptive allocation of attention in associative recognition toward informative (associative) features and away from uninformative (item) features.

This chapter discusses key features of computational models of event memory, also called “episodic” memory. Models aim to capture the representations and processes that enable us to perform a variety of episodic memory tasks, including recognition and free, cued, and serial recall. We review different ways in which models distinguish between the content and context of events; how different models represent event memories, including networks and vectors; and the types of processes that operate on memory representations to accomplish retrieval goals, including matching and search. Finally, we discuss how modeling serves broader scientific goals and can help bridge levels of explanation between the cognitive processes involved in memory and their neural instantiation.

A monotonic state-trace implies that a single latent factor is sufficient to explain the joint variation between two outcome variables across a set of conditions. There are, however, few methods available for assessing how much evidence a sample of data provides about whether the variables are truly monotonically related or not. We present a model that allows researchers to estimate the statistic M̂ which reflects the amount of evidence a dataset provides about whether two outcome variables are jointly monotonically related. This model is based on modeling the covariation between outcome measures in terms of a kernel function, which allows for computation of the latent derivatives of each outcome variable with respect to the other. M̂ is the posterior odds that these derivatives are all of the same sign and are thus monotonic. Simulations show that M̂ discriminates between monotonic and non-monotonic state traces and an example illustrates how the model can be applied to both continuous and binomial data from between-subjects, within-subjects, or mixed designs.