Purchase this article with an account.
Jared Peterson, Greg Erickson, Alicia Johnson, Jeff Dendurent, Lester Loschky; The Effects of Blur on Selective Visual Attention. Journal of Vision 2015;15(12):1071. doi: 10.1167/15.12.1071.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Dual task evidence suggests that visual blur is processed preattentively (Loschky et al., 2014). Interestingly, blur might be the first preattentive feature found to repel attention. Several experiments tested both of these hypotheses, using an L and T visual search task, and manipulating the presence of blur versus clarity. A pilot study found a level of blur that was clearly perceptible, but did not affect response times or accuracy compared to clear letters, thus confirming that both blurred and clear letters were equally legible. Experiment 1 compared RT for T targets with either all or no letters blurred, or the target or a single distractor as a blurred or clear singleton. Results suggested that blur did not repel visual attention, as shown by blurred singleton targets not having significantly longer RTs than the all-blurred and all-clear conditions. However, clarity seemed to capture attention—clear singleton targets had the fastest RTs, and clear singleton distractors near the target produced the longest RTs. Experiment 2 aimed to increase the effects of the blur manipulation by varying set size, increasing letter eccentricity, and increasing task difficulty by making L distractors appear more T-like (Jiang & Chun, 2001). With set size = 8, the results were consistent with blur not repelling visual attention (blurred singleton target = all-blurred & all-clear), while the clear singleton targets had the shortest RTs, consistent with clarity capture. Surprisingly, however, when there was less attentional demand (set size = 4), both clear and blurred singleton targets showed the shortest RTs, although capture by blur must be replicated. These visual search findings provide converging evidence to eye-tracking studies showing that when freely viewing scenes, observers typically attend to clear regions before blurred regions (Enns & MacDonald, 2012; Kahn, Dinet, & Konik, 2011; Loschky & McConkie, 2002; Smith & Tadmor, 2012).
Meeting abstract presented at VSS 2015
This PDF is available to Subscribers Only