Identifying relationships between cognitive processes across tasks, contexts, and time

Abstract

It is commonly assumed that a specific testing occasion (task, design, procedure, etc.) provides insights that generalize beyond that occasion. This assumption is infrequently carefully tested in data. We develop a statistically principled method to directly estimate the correlation between latent components of cognitive processing across tasks, contexts, and time. This method simultaneously estimates individual-participant parameters of a cognitive model at each testing occasion, group-level parameters representing across-participant parameter averages and variances, and across-task correlations. The approach provides a natural way to "borrow" strength across testing occasions, which can increase the precision of parameter estimates across all testing occasions. Two example applications demonstrate that the method is practical in standard designs. The examples, and a simulation study, also provide evidence about the reliability and validity of parameter estimates from the linear ballistic accumulator model. We conclude by highlighting the potential of the parameter-correlation method to provide an "assumption-light" tool for estimating the relatedness of cognitive processes across tasks, contexts, and time.

Speed-accuracy tradeoffs in decision making: Perception shifts and goal activation bias decision thresholds.

“…We also applied the SAT model to the perceptual discrimination task to directly test the hypothesis that the store pricing task activates goals in the subsequently experienced perceptual judgment task. We did this with a so-called “joint model” in that the two tasks were modeled in parallel in the same framework, which allows us to examine parameter associations across tasks (e.g., Wall et al, 2021). For the perceptual discrimination task, we assumed that the psychophysical representation of the perceptual stimulus was mapped to evidence accumulation rates in the SAT model via a cumulative normal link function (for similar approach, see Vandekerckhove et al, 2008).…”

Disentangling interacting psychological processes: A model-based approach to cognitive aging.

“…Fortunately, the race model framework used here is increasingly applied to a range of complex tasks involving instrumental learning, prospective memory, task switching, and decision conflicts [47][48][49][50] , and new cognitive-modelbased analyses have been developed 51,52 that address earlier concerns about how to rigorously measure capacity limitations 53 . New versions of correlation-based approaches are also being explored that can link cognitive-model parameters from different tasks through higher-order cognitive constructs 54,55 . This will afford researchers insights from individual-difference analyses while maintaining the reliability and validity advantages of cognitive models.…”

Efficient selection between hierarchical cognitive models: Cross-validation with variational Bayes.

“…We focus on the linear ballistic accumulator (LBA; Brown & Heathcote, 2008), which is simpler than many other EAMs in that it assumes no competition between alternatives (Brown & Heathcote, 2005), no passive decay of evidence (Usher & McClelland, 2001) and no within-trial variability (Ratcliff, 1978;Stone, 1960). This simplicity permits closed-form expressions for the likelihood function for the model parameters, which supports advanced statistical techniques including Bayesian methods based on MCMC and particle algorithms (Gunawan et al, 2020;Tran et al, 2021;Turner et al, 2013;Wall et al, 2021).…”