Language Familiarity and the Influence of Reading on Visual Search Strategies

Kelly A. Schwartz, Syracuse University

Abstract

Prior research indicates that subjects adopt a particular search pattern during visual search tasks. These search strategies are influenced by a variety of factors including color, orientation, size, and distinctness of stimuli. This project attempts to assess the level of influence language fluency has on visual search tasks. Experiments 1, 2, and 3 assess whether visual search tasks are affected by prior learned reading behaviors associated with particular languages. For these experiments, subjects were chosen based on fluency with Hebrew and English. These experiments used targets and distractors composed entirely of English, Hebrew, or symbols in different blocks. The experiments differed in the layout of the stimuli ranging from grid format of single characters and words, to text like formats using words. Language fluency groups were assessed on biases in search patterns using an eye tracker: left-to-right saccades during a block of English stimuli and right-to left saccades during a block of Hebrew stimuli. Results indicate a strong influence of prior learned reading behaviors on eye movements during visual searches. Subjects displayed differential eye movement patterns based on the stimuli for all three experiments. Subjects fluent in Hebrew showed eye search patterns similar to patterns observed while reading Hebrew text during the Hebrew block. During the English block, all subjects showed eye patterns similar to reading behaviors of English text. The fourth experiment determined differences in reaction times in a variety of conditions manipulating the language of the target word and the language of the distractor words. Subjects for this experiment were chosen based on their familiarity with the languages (Spanish and English) of the stimulus words. The results from this experiment indicate a reaction time cost for subjects fluent in both languages of the stimuli. More attention is needed to distinguish reaction time costs for bilinguals, but this finding is consistent with past research. This experiment failed to find a clear effect of language fluency on search times for Spanish words, nor did we find strong evidence of a linguistic pop-out effect. Collectively, these results provide valuable insight into the interaction of reading and visual search behavior.