Abstract
Mack and Rock (1999) had argued that attention is needed for perception through a phenomenon called inattentional blindness (IB) and that the standard gestalt grouping is not automatic but, requires attentional resources. The current investigation was motivated to assess the role of attention in computation of summary statistics of size estimation. Eighty naïve observers (Mage=26 years) participated in two IB experiments: Exp1(N=40; mean size estimation) and Exp2 (N=40; member size estimation). Each participant ran four trials, tasked to report the longer arm of the fixation cross (100ms)- either vertical or horizontal. On the fourth trial, unbeknownst to the participant, a set of four circles flashed simultaneously around the fixation cross. Then they were first asked (Q1) whether they saw anything else along with the fixation cross on the screen that was not there in previous trials. Then they were given two circles and asked (Q2) which of the two was the mean size of the circle set (Exp1) or which one of the two circles was present in last trial. 33.3% observers failed to see the set of circles in the IB trial in the mean estimation task and 29.0% in the member task. Of these, 72.7% accurately reported the mean size whereas 66.6% accurately reported the member size. Significant difference was observed between the expected and the observed value (χ2 =4.55; p=0.03) in the mean estimation but not for the member task (χ2 =0.73; p=0.39). Further, no significant difference was observed between the two tasks. For trials in which they do report consciously perceiving something (Q1=Yes; divided attention), 72.7% observers were accurate in mean task compared to only 59.9% in member task. Performance in the mean and member tasks was similar across conditions. Overall results indicate that estimation of summary size statistic also does require some attention.