A Modeling Pipeline for Inferring Hidden Rhythms in Behavioral Data: Application to the Death Implicit Association...

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Psychometrics

Background:

• Traditional behavioral analyses often rely on average response time (RT) and accuracy, which may obscure nuanced individual differences.
• Cognitive flexibility tasks, like the Death Implicit Association Test (DIAT), present challenges where average metrics might not fully capture behavioral complexity.

Purpose of the Study:

• To demonstrate that temporal patterns in behavioral data, beyond average RT, can provide meaningful discriminative power.
• To introduce a modeling pipeline for inferring temporal and dynamical patterns from RT data.
• To explore the utility of these patterns in identifying characteristics associated with suicidal ideation.

Main Methods:

• Utilized a modified Death Implicit Association Test (DIAT), termed the Brief Death Implicit Association Test (B-DIAT).
• Developed and applied a novel modeling pipeline to extract temporal and dynamical patterns from RT data across task trials.
• Correlated extracted behavioral patterns with suicidal ideation scores (SI).

Main Results:

• Average RT per condition in the B-DIAT did not effectively distinguish between participant traits.
• Extracted temporal patterns from RT data revealed unique behavioral characteristics.
• These temporal patterns showed a significant association with suicidal ideation, as indicated by SI scores.

Conclusions:

• Temporal dynamics in behavioral data offer richer insights than average metrics alone.
• The developed modeling pipeline can effectively infer discriminative behavioral patterns from RT data.
• This approach has potential applications in understanding cognitive mechanisms and behavioral deviations across various tasks, including mental health assessments.