The Cost of Regulating Effort: Reward and Difficulty Cues With Longer Prediction Horizons Have a Stronger Impact on Performance

Kukkonen, Nanne; Braem, Senne; Allaert, Jens; Eayrs, Joshua O.; Prutean, Nicoleta; Steendam, S. Tabitha; Boehler, C. Nico; Wiersema, Jan R.; Notebaert, Wim; Krebs, Ruth M. · 2025 · OpenAlex-citations

DOI: 10.5334/joc.415

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the cognitive costs associated with dynamically regulating mental effort, challenging the assumption that individuals can flexibly adjust effort expenditure without penalty. While existing neuroeconomic models posit that effort is optimized based on expected rewards and task difficulty, few studies have examined whether the act of regulation itself incurs a cost. The authors hypothesized that frequent, trial-by-trial adjustments to effort are costly, leading participants to ignore cues with short prediction horizons. Conversely, cues valid for multiple consecutive trials (long prediction horizons) should allow for more effective effort regulation, as the cost of adjustment is amortized over time. To test this, the researchers conducted four experiments using two distinct tasks: a random-dot-motion (RDM) perceptual decision-making task (Experiments 1 and 2) and a color-naming Stroop cognitive control task (Experiments 3 and 4). In each task, participants received cues indicating the reward magnitude (high vs. low) and task difficulty (easy vs. hard) for upcoming trials. The critical manipulation was the prediction horizon: in Experiments 1 and 3, cues applied only to the immediate next trial (short horizon), whereas in Experiments 2 and 4, cues remained valid for a block of six consecutive trials (long horizon). The RDM task also included intermediate difficulty trials to isolate the effect of difficulty expectation from actual difficulty. Performance was measured via reaction times and accuracy, analyzed using Generalized Linear Mixed Models. The results demonstrated that participants effectively utilized cue information only when it applied to multiple trials. In the RDM task, high reward expectancy improved accuracy, particularly on easy trials, but this facilitative effect was significant only in Experiment 2 (long prediction horizon), not in Experiment 1 (short horizon). Similarly, in the Stroop task, the reward-related facilitation of reaction times was observed exclusively after cues with a long prediction horizon. When cues changed every trial, participants largely ignored the incentive information, suggesting that the cost of constantly recalibrating effort outweighed the potential benefits. These findings indicate that effort regulation is not cost-free; rather, it involves a trade-off between the benefits of optimized performance and the cognitive cost of frequent adjustment. The study concludes that lower adjustment frequency, enabled by longer prediction horizons, compensates for this regulatory cost. This challenges standard models of effort allocation that assume dynamic, trial-wise optimization, suggesting instead that humans adopt a more stable effort strategy to avoid the aversive cost of constant regulation. The results highlight the importance of prediction horizon in understanding how incentives influence cognitive control and performance.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-25; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• mental demand