Abstract
People can make judgments about statistical properties (mean, variance) from object features. There is a claim that ensemble judgments do not require attentional resources. In the study by Bronfman et al. (2014), participants could judge the color diversity of letters briefly presented in a 4-by-6 grid while memorizing the letters in a cued row, and the presence/absence of ensemble judgments did not influence the letter memory task performance. Here we tested a plausible alternative explanation to this result: participants might have judged the color diversity using several letters around the cued row where they paid attention. In the present study, participants viewed letters in a 5-by-5 grid and memorized the letters in one of the five rows cued before the array presentation. After the array disappeared, participants reported the letter at one of the locations in the cued row. Then they judged the diversity of font weights of all letters (high vs. low). In Experiment 1, we manipulated the font-weight diversity for the two rows adjacent to the cued row and the other two rows separately. We found that diversity judgments were more affected by the letters in the adjacent rows. In Experiment 2, we manipulated the distance between the letters in the cued row and the letters in the other rows, by adding gaps above and below the cued row. We found that diversity judgments were less accurate when there were gaps around the cued row, that is, when fewer letters fall within the attentional window centered on the cued row. Our results suggest that letters around the cued row (i.e., letters within the window of focused attention for the primary task) can contribute to ensemble judgments and explain how additional ensemble judgments could have cost no additional attentional resources.