A general adaptive deadline algorithm for calibrating cognitive task difficulty

Effort-based decision-making is a highly studied topic that investigates how agents allocate resources to exert physical or mental effort to overcome task difficulty and obtain a reward. However, designing experiments investigating effort-based decision-making is complex, as the experimenter needs to ensure that all participants experience the same task difficulty, regardless of their prior proficiency in the task. While effective methods for solving this problem have been proposed for physical tasks, participant-specific calibration of task difficulty in a mental effort task remains challenging, above all if the calibration procedure needs to be implemented in a short time window to minimize the overall duration of the experiment. In this manuscript, we propose a novel and computationally lightweight method, named Fast Difficulty Calibration algorithm, that allows for the fast calibration of task difficulty by dynamically adapting the response time limit during trial execution. We test this method on a mental arithmetic task with four difficulty levels, and show that only 20 calibration trials are sufficient to make the subjective task difficulty similar for all the participants.