Dual and Single Route Models for Beginning Readers A Comparison by Means of Multinomial Processing Tree Models

There has been a lot of attention for the idea that the reading of a single word (visual word recognition) involves a single mechanism only. This mechanism first maps the orthographic input onto a sublexical phonological code via which, in a second step, the lexicon is accessed. This mechanism is called a single route phonological model, and it should be contrasted with a dual route model, which also assumes an orthographic route. This orthographic route maps the orthographic input onto a lexical orthographic code without phonological recoding. In this paper, both the single route phonological and the dual route models were formulated as multinomial processing tree (MPT) models. These two MPT models were applied to the data of two experiments in which the participants (children in Grades 1 and 2) had to give a combined naming and lexical decision response to four types of stimuli (words and three types of nonwords). The dual route model gave a much better explanation of these data than the single route phonological model