Does superior visual working memory capacity enable greater distractor suppression?

ABSTRACT

We asked whether individuals high in working memory capacity have a superior ability to proactively suppress features. If so, it would help explain why these individuals are more resistant to attention capture. We tested this hypothesis using the capture-probe paradigm employed in Lien et al. (2022. On preventing attention capture: Is singleton suppression actually singleton suppression? Psychological Research, 86(6), 1958–1971). Participants (N = 112) performed a colour change detection task, assessing visual working memory capacity. They then performed a visual search task (70% of the trials) intermixed with probe tasks (30% of the trials). For the visual search task, either a salient colour singleton distractor or non-salient distractor (a triplet) appeared with the target object. For the probe recall task, participants reported probe letters that briefly appeared inside each object. Replicating Lien et al., a suppression effect on probe recall accuracy was observed for both salient singletons and non-salient triplets. Critically, high and low visual working memory capacity individuals showed statistically equivalent ability to suppress colour distractors. These findings suggest that proactive suppression is not the mechanism by which high-capacity individuals achieve greater resistance to capture. Proactive suppression may be an implicit process that does not require special working memory capabilities.

KEYWORDS:

• Attention capture
• suppression
• visual working memory capacity
• visual search

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Gaspar et al. (2016) divided their 48 participants into high-K, medium-K, and low-K groups, and observed the absence of a P_D in the low-K group only. To provide a finer-grained analysis, we followed Gaspar et al.’s approach and sorted participants based on their K-value into 3 groups (high-K [N = 37], medium-K [N = 38], and low-K [N = 37]). A significant suppression effect was still observed for the low-K group (−5%±2%), |t(36)| = 6.84, p < .001, dz = 1.12. This effect was not significantly different from either the high-K or medium-K group (-7%±2% and −6%±2%, respectively), |ts| < 1.0.

2 We also used JASP to run a linear mixed effect analysis of the relationship between the suppression effect, K-value, and trial type (singleton vs. triplet; dummy-coded). As fixed effects, we entered K-value and trial type into the model. We found no significant effects of K-value ($\widehat{\text{β}}$ = −.004, p = .48) or trial type ($\widehat{\text{β}}$ = .01, p = .10). Furthermore, no significant interaction was observed between K-value and trial type ($\widehat{\text{β}}$ = .003, p = .44).