Quantifying mechanisms of cognition with an experiment and modeling ecosystem

Abstract

Although there have been major strides toward uncovering the neurobehavioral mechanisms involved in cognitive functions like memory and decision making, methods for measuring behavior and accessing latent processes through computational means remain limited. To this end, we have created SUPREME (Sensing to Understanding and Prediction Realized via an Experiment and Modeling Ecosystem): a toolbox for comprehensive cognitive assessment, provided by a combination of construct-targeted tasks and corresponding computational models. SUPREME includes four tasks, each developed symbiotically with a mechanistic model, which together provide quantified assessments of perception, cognitive control, declarative memory, reward valuation, and frustrative nonreward. In this study, we provide validation analyses for each task using two sessions of data from a cohort of cognitively normal participants (N = 65). Measures of test-retest reliability (r: 0.58-0.75), stability of individual differences (ρ: 0.56-0.70), and internal consistency (α: 0.80-0.86) support the validity of our tasks. After fitting the models to data from individual subjects, we demonstrate each model's ability to capture observed patterns of behavioral results across task conditions. Our computational approaches allow us to decompose behavior into cognitively interpretable subprocesses, which we can compare both within and between participants. We discuss potential future applications of SUPREME, including clinical assessments, longitudinal tracking of cognitive functions, and insight into compensatory mechanisms.

“The eyes are the window to the representation”: Linking gaze to memory precision and decision weights in object discrimination tasks.

“…Here, we conducted analyses for identifying subgroups of participants who demonstrate these key behaviors. Because Deng and Sloutsky’s key effects pertain to aggregate group-level behaviors rather than individual subjects, we first classify participants into groups using individual-level, model-based cognitive assessment techniques (Weichart et al, 2021; Wiecki et al, 2015). We then conduct comparisons between groups to verify that the contrasting behavioral correlates of attention described by Deng and Sloutsky (2016) are indeed observable within the current participant pool, despite controlling for age.…”

“…We applied a suite of GCM variants (Nosofsky, 1986) with separate freely estimated distributions of α for recognition and categorization and used a switchboard analysis to characterize individual-level attention (Turner et al, 2018). Our approach follows from relatively recent efforts in model-based cognitive assessment, in which well-established cognitive models are used to delineate participants according to the latent mechanisms that plausibly underlie their behaviors (Darby & Sederberg, 2022; Weichart & Sederberg, 2021; Weichart et al, 2021). Here, the relevant mechanism for delineation is attention, and the four variants of interest are summarized in Table 2.…”

Transparency, replicability, and discovery in cognitive aging research: A computational modeling approach.

“…Each block of the task contained 48 trials (four trials for each of the 12 conditions listed in Table 1) and lasted approximately 2.5 min. In addition to the CAR task, all participants completed other tasks from a larger cognitive battery (Weichart et al, 2021), which we do not address in the current manuscript. At least one intervening task was presented between each block of the CAR task.…”

“…Models are most often employed for the purpose of developing and testing mechanistic theories of cognitive processes in healthy, and typically young, adults (Weichart et al, 2021). Increasingly, however, modeling frameworks have been applied to jointly test theories and gain insight into mechanistic differences between individuals from distinct groups.…”