The dilemma of experience in voluntary attention control: How choice clarifies a common mechanism across opposing approaches.

This article attempts to clarify the endogenous mode of Ray Klein's three-dimensional taxonomy of attention by addressing a central dilemma in measuring voluntary attention control (VAC): how to handle the contamination of response time (RT) metrics by task experience. The performance-centred and autonomy-centred approaches represent diametrically opposed responses to this dilemma. The performance-centred approach uses uninformative cue contexts to isolate "pure" VAC, arguing that informative cues contaminate RT metrics via associative learning of cue-target probabilities. Conversely, the autonomy-centred approach argues that authentic volition requires self-generated choices in informative contexts-where the environment does not conflict with intentions-and utilizes a mediation model to separate VAC from contaminated RT measures. Rather than attempting to evade the effects of experience, the present study focused directly on them. We distinguished between two distinct experience-based effects: a practice-based effect tied to action frequency and a probability-based effect tied to cue-target statistics. We then examined the extent to which these effects depend on self-generated intentions by forcing participants to change cues at preset rates (0.25 vs. 0.75). Results indicated that the practice-based effect was weakened when choices were forced (as opposed to free), and the probability-based effect became undetectable. Both effects were significantly stronger when cue changes were self-generated and freely chosen. Altogether, the present findings underscore the dilemma of experience in VAC and highlight the important role of self-generated intentions in driving both experience-based RT effects and VAC. (PsycInfo Database Record (c) 2026 APA, all rights reserved).