Factor analysis of 57 variables in visual word recognition

Based on current models in visual word recognition 57 variables were selected that influence performance in lexical decision and naming tasks and are connected to effects of frequency, neighborhood and consistency on the word, subcomponent and letter level. Two consecutive factor analyses using these variables and 1175 German monosyllabic four to six letter words yielded six orthogonal factors. The factors are interpreted as (F 1) a factor of general word-likeliness, (F2) a factor for the middle and (173) one for the initial part of the word, (F4) feedback consistency of the end of the word, (F5) frequency-, and neighborhood of the end of the word, and (176) feedforward consistency of the end of the word. Four regression analyses on previously collected data from naming and lexical decision tasks show that the factors explain a substantial amount of variance in response times. This theoretically guided factor analytical approach can provide both a useful tool for researchers not primarily concerned with visual word recognition but relying on verbal material and a framework for an approach to visual word recognition based on invariant stimulus structure.