Temporal dynamics of motivation-cognitive control interactions revealed by high-resolution pupillometry

Abstract

Motivational manipulations, such as the presence of performance-contingent reward incentives, can have substantial influences on cognitive control. Previous evidence suggests that reward incentives may enhance cognitive performance specifically through increased preparatory, or proactive, control processes. The present study examined reward influences on cognitive control dynamics in the AX-Continuous Performance Task (AX-CPT), using high-resolution pupillometry. In the AX-CPT, contextual cues must be actively maintained over a delay in order to appropriately respond to ambiguous target probes. A key feature of the task is that it permits dissociable characterization of preparatory, proactive control processes (i.e., utilization of context) and reactive control processes (i.e., target-evoked interference resolution). Task performance profiles suggested that reward incentives enhanced proactive control (context utilization). Critically, pupil dilation was also increased on reward incentive trials during context maintenance periods, suggesting trial-specific shifts in proactive control, particularly when context cues indicated the need to overcome the dominant target response bias. Reward incentives had both transient (i.e., trial-by-trial) and sustained (i.e., block-based) effects on pupil dilation, which may reflect distinct underlying processes. The transient pupillary effects were present even when comparing against trials matched in task performance, suggesting a unique motivational influence of reward incentives. These results suggest that pupillometry may be a useful technique for investigating reward motivational signals and their dynamic influence on cognitive control.

Keywords: cognitive control; incentive; motivation; pupillometry; reward.

Figure 1

Trial structure with timing. Example AX (target) trial, with both non-incentive (green square) and incentive (green dollar sign) precues shown. One of these two precues preceded each contextual cue-probe letter pair and, in the reward block, indicated the presence/absence of incentive value for the trial.

Figure 2

Block-related (baseline vs. non-incentive trials) and trial-type (non-incentive vs. incentive trials) effects on task performance: (A) with error rates as a dependent measure; (B) with RTs as a dependent measure.

Figure 3

(A) Pupil timecourses as a function of incentive status for the sustained incentive contrast (baseline vs. non-incentive trials, averaged across trial-types). (B) Sustained incentive effects (as averaged pupil magnitudes) at pre-trial period (−200 to 0 ms).

Figure 4

(A) Pupil trial timecourses as a function of incentive status and trial for the incentive cue contrast. (B) Incentive trial effects (as averaged pupil magnitudes) at pre-probe onset period (1950–2200 ms).

Figure 5

(A) Trial-evoked (normalized) pupil timecourses in baseline and non-incentive trials within the reward block. (B) Trial-evoked pupil activity (as averaged pupil magnitudes) in baseline and non-incentive conditions at pre-probe onset period (1950–2200 ms).

Figure 6

(A) Reward block pupil timecourses (correct AX trials only) split by incentive and reaction time (faster/slower than RT criterion) for each participant. (B) Incentive trial and RT effects at pre-probe onset period (1950–2200 ms). These data reveal that pupil dilation is associated with both task performance and incentive status, and that these effects do not appear to interact.

Figure A1

Task performance and pupil data as a function of 50-trial sections, for baseline and non-incentive trials in the reward block: (A) with error rates as a dependent measure; (B) with RTs as a dependent measure; (C) with pupil dilation at pre-trial period (−200 to 0 ms) as a dependent measure.