Statistical Learning of Target Location Probability in Children and Adults

Abstract

Learning environmental regularities allows us to make predictions and guide behavior. Growing evidence of location probability learning (LPL) has shown that the statistical regularity of target locations affects spatial attention allocation. However, past studies on LPL have mostly focused on adults’ learning. To achieve a comprehensive understanding of the mechanism of this learning, we investigated the effect of target location probability on 5- to 9-year-old children’s visual search in comparison with that of adults. Both children and adults responded faster when the target appeared in the high probability “rich” quadrant than in the low probability “sparse” quadrants of the search space. This attentional bias toward the rich quadrant persisted even when the target was equally likely to appear in all four quadrants. Importantly, the magnitude of the bias was constant across various ages of participants and did not depend on individual differences in executive functions. Taken together, these results provide novel and converging evidence that implicit statistical learning of target locations occurs early in development and remains stable until early adulthood, which is a distinct developmental pattern from explicit goal-driven spatial attention learning.

Age-related differences in the statistical learning of target selection and distractor suppression.

“…More in general, these results may suggest that automatic and implicit-based attentional learning mechanisms may be preserved even despite a more general deficit in attentional mechanisms and reduced cognitive control induced by aging or development and neurocognitive disease (Amer et al, 2016). In line with this notion, unimpaired spatial location probability learning has been demonstrated not only in older adults (Jiang, 2018;Jiang et al, 2016;Twedell et al, 2017) but also in patients with Parkinson's disease (Sisk et al, 2018), in children (Lee et al, 2020;Yang & Song, 2021), and in autistic spectrum disorder (Jiang, Capistrano, et al, 2013).…”

“…In particular, when the target of visual search appears in specific spatial locations with different frequency, its selection is facilitated when it appears in a high-probability region compared with a low-probability region (Di Caro & Della Libera, 2021; Ferrante et al, 2018; Geng & Behrmann, 2005; Jiang et al, 2015; Schapiro & Turk-Browne, 2015; Wang & Theeuwes, 2018a, 2018b). These types of experience-driven effects on attention emerge very rapidly from the onset of stimulus probability imbalances (Di Caro & Della Libera, 2021; Ferrante et al, 2018; Huang et al, 2021; Jiang, Swallow, et al, 2013) and persist well after the task contingencies have been eliminated (Di Caro & Della Libera, 2021; Ferrante et al, 2018; Lee et al, 2020). This is true not only immediately after the experimental session (Di Caro & Della Libera, 2021; Ferrante et al, 2018; Giménez-Fernández et al, 2022; Jiang, Swallow, et al, 2013) but also for at least 1 week after the acquisition (Jiang et al, 2014), suggesting a long-lasting change in attentional priority that is not a mere temporary consequence of current cumulative expectations.…”

Is probabilistic cuing of visual search an inflexible attentional habit? A meta-analytic review

“…Figure 1 shows a schematic illustration of the displays and design in a typical experiment. Despite this change in the spatial distribution of targets, participants usually continue to respond faster in the unbiased stage when the target is located in the previously rich quadrant (e.g., Ferrante et al, 2018;Lee et al, 2020). This result can no longer be explained in terms of repetition priming, because during the unbiased stage repetitions are equally likely to appear in all regions.…”

Is probabilistic cuing of visual search an inflexible attentional habit? A meta-analytic review

“…Figure 1 shows a schematic illustration of the displays and design in a typical experiment. Despite this change in the spatial distribution of targets, participants usually continue to respond faster in the unbiased stage when the target is located in the previously rich region (e.g., Ferrante et al, 2018;Lee et al, 2020). In these studies, reaction times are typically analyzed through an ANOVA that includes region (rich or sparse) and stage (biased, unbiased) as factors.…”