Visual search for letters and words in letter arrays: performance and eye movements
In visual search among complex stimuli, one of the most important issues is the problem of chunking. In our experiments, we studied visual search for letters in letter arrays, using a modified selective attention test. Observers searched for a prespecified letter in an array of random letters which included words. Target letters were either always or never embedded in those words. In the control condition, there were no words in the array. We discovered a dissociation of letter search efficiency (no statistically significant differences between the three conditions) and subjective representation of one's performance (drastic differences between the two experimental conditions). To understand this dissociation, we compared time-limited search for prespecified letters and search for words in the same letter arrays combining behavioral data and eyetracking. Search efficiency differed in the two conditions: letter search proved much easier than word search (mean search performance was 75% for letters and 46% for words). The pattern of fixations also differed: when searching for words, observers performed more fixations both between and within words. At the same time, almost all participants who searched for letters noticed words in letter strings. Thus, automatic word segregation is possible, but task-driven chunking is a resource-consuming operation.
attention, visual search