Children's use of context in reading.

A series of investigations are reported which explore children's strategic use of the linguistic context in reading. The investigations sought to determine whether better readers make better use of the linguistic context than poorer readers, through the use of a more efficient strategy. Previous research had shown that the performance of better readers was in fact superior on tasks requiring the use of context, and the researchers tended to conclude that the readers differed in the strategies they used. However, it is argued here that there was no valid evidence that the readers differed in the strategies they employed, as the researchers had not controlled for the confounding effects of the readers' differing knowledge. The first series of the author's investigations were concerned with the readers' use of the succeeding context in a cloze-type situation. There were five experiments and an analysis of errors, in all involving 440 children aged seven, eight, and nine years. A test was specially constructed which would control for the readers' differing knowledge, and it was found that, on average, the better readers did make better use of the succeeding context, after controlling for their superior knowledge. The children's self-corrective behaviour was interpreted to suggest the following: that a minority of children made very poor use of the succeeding context, and that these children tended to be poorer readers; and that the majority of children made equally good use of the preceding and succeeding context. Another experiment was carried out (with 28 seven and eight year old children), and this was concerned with the readers' use of the context when reading aloud connected prose. It was concluded that the contextual acceptability of oral reading errors was not a reliable guide to the readers' use of context, partly because their acceptability is a function of the readers' decoding skills.