Abstract
Past research suggests task-relevant objects are better encoded into visual long-term memory than task-irrelevant objects. McClellan, Varakin, Renfro, and Hays (2018, VSS) demonstrated that the number of task-relevant object categories during study moderates this effect. When participants counted or memorized one or two object categories, task-relevant objects were better recognized on a subsequent memory test than task-irrelevant objects. However, when participants counted or memorized five object categories, the memory advantage for task-relevant over task-irrelevant objects was not significant. The current experiment replicates and extends these findings. Participants (N = 194) first viewed a series of objects while performing a cover task. Whereas our past participants either counted or memorized objects, participants in the current study pressed a button whenever task-relevant objects appeared. The number of task relevant object categories was one, two, three, or five. As in past work, participants completed a yes/ no recognition memory test that consisted of one task-relevant category and one task-irrelevant category. Similar to the results of McClellan et al. (2018), with recognition sensitivity as the dependent variable, there was an interaction between task-relevance and the number of relevant categories at study (p < .05). When one or two categories were relevant at study, the effect size of task-relevance was about d = 0.77 (ps < .001). When three or five categories were task relevant, effect sizes were smaller, d = .42 and d =.49 (ps = .005 and .002), respectively. These results reinforce the idea that limits of attentional selection and working memory influence encoding into VLTM. However, the results do not align with the idea that the limit is 3 or 4 items. First, the effect of task-relevance decreased between two and three task-relevant categories. Second, though diminished, there was still a statistically significant effect of task-relevance on VLTM when five categories were task-relevant during study.