Dynamic computational phenotyping of human cognition

Abstract

Computational phenotyping has emerged as a powerful tool for characterizing individual variability across a variety of cognitive domains. An individual’s computational phenotype is defined as a set of mechanistically interpretable parameters obtained from fitting computational models to behavioural data. However, the interpretation of these parameters hinges critically on their psychometric properties, which are rarely studied. To identify the sources governing the temporal variability of the computational phenotype, we carried out a 12-week longitudinal study using a battery of seven tasks that measure aspects of human learning, memory, perception and decision making. To examine the influence of state effects, each week, participants provided reports tracking their mood, habits and daily activities. We developed a dynamic computational phenotyping framework, which allowed us to tease apart the time-varying effects of practice and internal states such as affective valence and arousal. Our results show that many phenotype dimensions covary with practice and affective factors, indicating that what appears to be unreliability may reflect previously unmeasured structure. These results support a fundamentally dynamic understanding of cognitive variability within an individual.

Adaptive Integration of Perceptual and Reward Information in an Uncertain World

“…However, applying our approach to clinical populations or studies interested in individual differences would particularly benefit from more stable estimates. Among many factors that impact reliability, Schurr et al (2024) identify random noise in behavioral measurements that could arise from discrepancies given the current application of the slider. Improvements to the slider design, including cues about previous estimates (to reduce motor noise) and modifying the starting point of the slider (requiring fewer adjustments), could increase the overall reliability of the parameters.…”

Reliable, rapid, and remote measurement of metacognitive bias

“…This was in line with the task design, as stability in task performance is maintained by adjusting in task difficulty, which increased across sessions. In addition to practice factors, affective state can be a source of significant temporal variability in cognitive task performance 35 . However, we did not measure any changes in affective state or psychiatric symptoms across time.…”

“…Alternatively, a less pessimistic view of poor reliability among cognitive outcomes is that behavioural tasks provide legitimate estimates of momentary cognitive capacities when tested, but these are liable to fluctuations over time 34 . In line with this, cognitive capacities tend to temporally covary with affect and practice factors, which are often not accounted for in reliability assessments 35 . This further illustrates the need for within-subject longitudinal assessments in computational psychiatry, to infer individual clinical and cognitive phenotypes.…”

Attention and learning strategies reveal distinct profiles of psychiatric traits

“…This suggests that the practice effects of computational psychiatry assays offer rich set of additional phenotypic information. Indeed, a recent study utilizing a battery of seven behavioral tasks over twelve weekly sessions found both pronounced practice affects, and that affective state significantly impacted on task behavior (Schurr et al, 2024).…”