On the relation between single word and multiple word processing during learning to read

Abstract

The current study investigated the development of single word processing and multiple word processing skills in French-speaking children from Grade 2 to Grade 6. A total of 150 children participated in two tasks: a Lexical Decision Task (LDT) and a Grammatical Decision Task (GDT). The LDT was used to test single word processing abilities, whereas the GDT was used to test multiple word processing abilities, with stimuli presented at varying display times (LDT: 83–300 ms; GDT: 150–700 ms). Signal detection theory analysis revealed that all children performed well in the LDT, whereas only Grade 4 and Grade 6 children performed above chance in the GDT. A cluster analysis was used to investigate the different types of relation between sensitivity (d’) in the LDT and GDT. The analysis revealed two clusters that differed in reading fluency and sensitivity in both tasks. Children from Cluster 2, who exhibited the highest sensitivity in the LDT, were the only ones to perform on average above the chance level in the GDT. Moreover, a strong correlation (r = .64) between LDT and GDT performance was found in this group. Finally, we found that a sensitivity of 1.95 in the LDT almost perfectly predicted cluster membership. Such sensitivity was achieved at Grade 3, suggesting that the ability to process multiple word sequences first requires sufficient efficiency in processing words in isolation. Once this turning point is reached, single word processing skills support the development of multiple word processing, which could take several years to mature fully.