Can post-error dynamics explain sequential reaction time patterns?

Area of Science:

• Cognitive Psychology
• Human-Computer Interaction
• Neuroscience

Background:

• Human performance in sequential tasks exhibits systematic dependencies on preceding stimuli.
• Reaction times (RTs) and error rates are influenced by trial history.

Purpose of the Study:

• To investigate human error dynamics in sequential two-alternative choice tasks.
• To analyze sequential effects on RTs, differentiating between error and correct responses.
• To identify and model the sequential RT tradeoff.

Main Methods:

• Analysis of RTs and error rates across sequences of stimuli.
• Utilizing data from choice tasks with first-order Markov process stimulus generation.
• Reanalysis of prior data and acquisition of new experimental data.
• Application of a pure drift diffusion model with sequential updates.

Main Results:

• A sequential RT tradeoff was identified: fast RTs on error trials correspond to slow RTs on correct trials.
• Stimulus sequence properties, particularly alternation probability, significantly influence RTs.
• Existing models failed to capture these observed relationships.
• Sequential updates to drift diffusion model parameters accounted for RT trends.

Conclusions:

• Error-based parameter adjustments are crucial for accurately modeling sequential effects in human performance.
• Understanding human error dynamics provides insights into cognitive processes.
• The identified RT tradeoff offers a novel perspective on decision-making under sequential constraints.