The effects of reading goal and textual support on the loci of children\u27s inferential generation

This study addressed the question of whether inferences required for textual comprehension are drawn at encoding, for both children and adults. Third-, fifth-, and eighth-grade students, and adults read a series of four-sentence stories where a consequence was strongly implied. The sentence implying this consequence was either presented in the last sentence of the story (a forward inference), was followed by a comprehension-inducing sentence (a bridging inference), or was followed by a sentence which explicitly mentioned the consequence (a bridging-explicit inference). In addition, one-half of the subjects engaged in an integrative reading task, while the other half engaged in more superficial processing. Whether or not the consequence inferences were drawn at encoding was assessed through the use of reading times to the final sentences in the bridging and bridging-explicit versions. In addition, question-answering latencies and error percentages to questions probing the inference were also obtained. The results supported the main hypothesis that the bridging inferences would be generated during reading (encoding) while the forward inferences would not be generated until test time (retrieval). Although reading times and question-answering latencies decreased with each higher grade level, the pattern of results was identical for all four age groups. This suggests that children as young as the third grade have adult-like inferential processing capabilities. The reading goal manipulation was ineffective in prompting additional inferential processing, for all four age levels. The only interpretable effect of reading condition was that more developmental improvement was noted in the integrative reading condition than in the superficial reading condition. Post-hoc analyses suggested that the younger readers were somehow preparing themselves in advance to perform the integrative reading task, and this advance preparation taxed their available attentional resources. It was argued that further research needs to focus on finding a reading goal manipulation that is successful in altering young children\u27s reading comprehension