Evaluating Response Times in Letter Strings as Crowding Magnitude Increases

Background: Individuals with severe central vision loss usually adopt eccentric viewing for tasks such as reading. While crowding has been implicated as a dominant cause for slower reading speeds with eccentric viewing by reducing letter recognition accuracy, it is possible that crowding could also impose pre-lexical processing delays, which may also contribute to slower reading speeds with increasing viewing eccentricity. Therefore, the purpose of this study is to measure the response times to letter strings with increasing crowding magnitude at different eccentricities to explore pre-lexical processing delays. Methods: Letter recognition accuracy and response times were measured in 11 subjects with normal vision for the central letter presented within a trigram. The trigrams were presented for 100ms at a fixation distance of 57 (114) cm. The trigrams were presented at three different locations; the central letter centered at fixation as well as 3 degrees to the left and right of fixation with varying degrees of crowding. All letters were presented with high contrast (0.84) lowercase Courier font. Stimulus eccentricity and inter-letter spacing were presented randomly within a single block of trials. Results: Letter recognition accuracy for central and peripheral presentations decreased with decreasing separation of flanking targets. Response times increased proportionally with decreasing inter-letter separation for all 3 viewing separations. In the case of isolated letters, response times were generally slower in the left hemifield compared to the right hemifield despite equivalent response accuracy. Conclusions: Visual crowding does indeed induce delays in perceptual processing time for the extraction of letter elements embedded in letter strings. Furthermore, this study demonstrates that perceptual processing delays induced by crowding occurs at pre-lexical processing stages, especially given that the letter strings employed in the current study lacked lexical information.