Among 270 adults wired to eye-tracking rigs in a joint University of Tehran and University of Calgary experiment, the difference between working for a reward and working for a choice collapsed into a single behavioral measurement: where the eyes went first. According to a new Frontiers study published this month, performance-contingent incentives — whether conspicuous or barely visible — reliably raised both time-on-task and accuracy on a demanding memory test, while simply offering a choice of how to do the work moved neither metric.
The 3×3 design
The lab experiment randomized participants across a three-by-three factorial matrix: three choice levels (no-choice, choice, control) crossed with three reward levels (control, non-salient, salient), yielding nine conditions. Motivation was operationalized as time spent on the memory task; performance as the number of correct answers. Eye-tracking captured gaze direction toward reward-related stimuli, allowing the team to test whether visual attention mediated the behavioral gains — a direct empirical probe of Motivational Congruence Theory's cognitive predictions in a controlling context.
Where the mechanism sits
Both reward conditions moved the outcome variables. Choice did not. The mediating channel was visual: gaze directed at the reward cue predicted longer engagement and higher accuracy on the test. The implication is mechanistic — reward structures capture attention first, and the downstream motivational and performance gains ride on that attentional lock. Autonomy, in a controlling context, added processing load without producing measurable output. The eye, not the option set, carried the effect.
What to recalibrate in practice
For clinicians, performance coaches, and learning designers running attention-heavy interventions — cognitive rehabilitation blocks, exam-prep modules, high-stakes training drills — the instruction is concrete. Calibrate the salience of the reward signal before expanding the choice architecture. Make the incentive visible, even subtly so, and let the gaze do the motivational lifting. Add autonomy only after the attentional lock is secured, and instrument the protocol with a measurable proxy for engagement — fixation counts, latency to first fixation, dwell time on the cue — not self-reported motivation. Decoration kills precision; a clear stake fixes the line of sight.