Attention capture during visual search: The consequences of distractor appeal, familiarity, and frequency

ABSTRACT

Visual distractions can lure our attention and impede our everyday performance, especially if they are highly meaningful and appealing to the observer. The current research assessed how semantically rich, personally relevant distractors (i.e., cartoon characters), either appealing or neutral, capture attention and whether the frequency with which we encounter these distractors can impact the effects. Participants were slower to identify a target letter in the presence of a neutral distractor relative to an appealing distractor, reflecting covert attentional capture. However, this effect reversed when appealing distractors appeared less frequently than neutral distractors. Collectively, the evidence suggests that the amount of capture observed overall likely depends on the interplay between a distractor’s semantic salience (i.e., the amount of meaningful knowledge an observer has about the distractor), its affective salience (i.e., how the observer feels about the distractor), and how frequently it is encountered.

KEYWORDS:

• Visual attention
• attention capture
• affect
• frequency
• eye movements

Acknowledgement

The authors have no financial or non-financial competing interests in this manuscript. The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, nor the U.S. Government. LT Biggs is a military service member. This work was prepared as part of his official duties. Title 17 U.S.C. §105 provides that ‘Copyright protection under this title is not available for any work of the United States Government.’ Title 17 U.S.C. §101 defines a U.S. Government work as a work prepared by a military service member or employee of the U.S. Government as part of that person’s official duties.

Disclosure statement

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

Notes

1 We later conducted four post-hoc power analyses using 80% power and effect sizes as found in four studies that used similar paradigms to the one used in the current work (Cohen’s d = 2.16, N = 16; Forster & Lavie, 2008a Experiment 1; F = 67.9, N = 14; Forster & Lavie, 2008b Experiment 3; F = 4.12, N = 18; Biggs et al., 2012 Experiment 2; F = 9.03; N = 22; Biggs et al., 2012 Experiment 3). To account for differences in reported values, we converted t-values and F-values (Lakens, 2013) for our sample size calculations (G*Power; Faul, Erdfelder, Lang, & Buchner, 2007). These calculations revealed four different target sample sizes: N = 4 (Forster & Lavie, 2008a Experiment 1), N = 4 (Forster & Lavie, 2008b Experiment 3), N = 22 (Biggs et al., 2012 Experiment 2), and N = 37 (Biggs et al., 2012 Experiment 3). The average of these four sample size calculations was 16.8 participants.

2 Recruitment and data collection occurred prior to the release of Pokémon Go, a popular augmented reality game that was released the summer of 2016. Therefore, participants were self-selected fans prior to Pokémon Go, which repopularized the Pokémon franchise.

3 The sample size for Experiment 2 was similar to that of Experiment 1. We later conducted four power analyses using 80% power and effect sizes as found in four published studies that used a similar disengagement paradigm as the one used here (F = 4.06, N = 22; Biggs et al., 2012 Experiment 4; F = 11.43, N = 40; Boot & Brockmole, 2010 Experiment 1; t = 2.40, N = 8; Brockmole & Boot, 2009 Experiment 2; t = 2.87, N = 20, Wright, Boot, & Brockmole, 2015). These calculations revealed four different target sample sizes: N = 45 (Biggs et al., 2012, Experiment 4), N = 30 (Boot & Brockmole, 2010 Experiment 1), N = 13 (Brockmole & Boot, 2009), and N = 22 (Wright, Boot, & Brockmole, 2015). The average of these four sample size calculations was N = 27.5, and, thus, we suggest that a more ideal sample size would have consisted of at least twenty-eight participants.

4 We later conducted a post-hoc power analyses using 80% power and the effect size as found in Experiment 1 (${\eta }_p^2$  = .244), which revealed an approximated target sample size of at least thirty-six participants in total. Therefore, we believe our sample size for Experiment 3 was appropriate.

5 Because demographic and opinion surveys were not administered, which was used as an exclusionary criteria in the previous two experiments, data from all participants were analyzed.