Purchase this article with an account.
Beth Stankevich, Arni Kristjánsson, Joy Geng; Reward- and space-based repetition priming is weighted by task relevance.. Journal of Vision 2015;15(12):1349. doi: 10.1167/15.12.1349.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Reward-associated features are prioritized during attentional selection. This priority is long-lasting even when task-irrelevant (Anderson et al., 2013) suggesting that reward associations are a particularly potent attentional cue that operates through selection history (Awh et al., 2012) and is separate from typical top-down and bottom-up cues (Posner, 1980). Reward-associated features have also been shown to magnify repetition priming effects (Kristjánsson et al., 2010). We investigated whether repetition priming for rewarded and non-rewarded dimensions would be modulated by top-down knowledge of current task goals. We employed a target discrimination task consisting of two phases, an initial rewarded phase (participants learned a color-reward mapping) followed by an extinction phase (learned color-reward mapping became invalid). Consistent with Kristjánsson et al., the repetition priming effect was bigger for rewarded than non-rewarded targets, but only during the rewarded phase. The novel finding was an interaction between repetition priming of the reward-associated color and target location: the priming effect was similar in size during the rewarded phase, but increased for location repetitions and decreased for reward-associated color repetitions in the extinction phase. This suggests that repetition priming is sensitive to the contextual relevance of task information. To ascertain whether the observed interaction was due to changes in reward-associated mapping or specific to changes in top-down knowledge of reward relevance, we performed a second study where the color-reward mapping reversed in alternating blocks. Priming for rewarded color and target location was comparable in all blocks, suggesting that simply changing reward mapping does not alter the magnitude of repetition priming for either information type. These results suggest that repetition priming is modulated by the contextual knowledge of task-relevant dimensions. The selection history of multiple features results in repetition priming effects, the magnitude of which is weighted according to task-relevance.
Meeting abstract presented at VSS 2015
This PDF is available to Subscribers Only