University of the Basque Country
When we read a word, not only the representation of that word but also the representations of its neighbors -all the orthographically similar words- are activated in our lexicon (Grainger et al., 1989). The neighborhood frequency effect is an example of this phenomenon, and it has been deeply studied in adults: words that have higher frequency neighbors are more difficult to identify than words that do not have neighbors of higher frequency (Carreiras, Grainger & Perea, 1997, Davis & Taft, 2005). This effect informs about the word coding mechanisms in the cognitive system of skilled readers. However, there is little evidence about when and how this effect emerges in children who are learning to read. We examined the neighborhood frequency effect in children of 2nd, 4th and 6th grade. In Experiment 1, children were required to name words that had both addition, deletion or transposition neighbors and control words that had no neighbors. In Experiment 2, a lexical decision experiment was run using the same stimuli. In both tasks, only children in 4nd and 6th grade were sensitive to neighborhood frequency, with addition neighbors leading to inhibitory effects. Interestingly, this effect was observed in reading times and % of errors in the naming task, but only in % of errors in the lexical decision task. We discuss these findings in terms of the progressive development towards a lexical reading strategy, and the different requirements of both tasks.