Purchase this article with an account.
E. M. Fine; The word-superiority effect does not depend on one's ability to identify the letter string as a word. Journal of Vision 2001;1(3):410. doi: 10.1167/1.3.410.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose. In the retinal periphery, the middle letters of 3-letter words are identified more accurately than the middle letters of trigrams, as predicted by the word-superiority effect (Fine; AAO, 1999). In the current study, the impact of lexical knowledge on this effect was investigated. Methods. Subjects identified the middle letters of 3-letter words and trigrams. On each trial, they simultaneously identified the letter and determined whether the stimulus they saw was a word or not. This was accomplished by presenting two response matrices, one labeled “words” and the other “nonwords”. Each response matrix consisted of the 26 letters of the alphabet. Following each stimulus presentation (150 msec), the response matrices were displayed until the subject responded by clicking on the target letter in the appropriate matrix. The stimuli were presented at 10 deg eccentricity in blocks either to the left or right of fixation or above or below fixation. Results. Subjects correctly identified the stimulus as a trigram on 74% of trigram trials and correctly identified the stimulus as a word on 66% of word trials. Overall, they correctly identified 50% of middle letters from the trigrams and 72% from the words; a 22% word-superiority effect. When subjects correctly identified the stimulus type, middle-letter accuracy was 65% for trigrams and 87% for words. When the stimulus type was identified incorrectly, middle-letter accuracy was 5% for trigrams and 41% for words. Conclusions. These data suggest that letters are more visible when they are part of a word than when they are part of a trigram, regardless of the observer's ability to identify the letter string as a word. Accuracy decreased for both stimulus types when the type was misidentified. However observers were able to identify the middle letters of words, while their ability to identify the middle letters of trigrams was at near chance levels. This finding suggests that whatever benefits accrue from the context in which letters are identified is, at least in part, independent of one's ability to recognize that context.
This PDF is available to Subscribers Only