The Effect of Illustrations on the Ability of Children to Draw Inferences While Reading Narrative Texts

Abstract

Materials used to teach reading in New Zealand schools, as elsewhere, generally include illustrations to help learners develop reading skills. The present research tests the claim that the availability of pictures can hinder rather than help the complex constructive processes of reading and considers whether or not boys are more at risk than girls. It does this by considering how pictures affect the ability of learners to practice making inferences and how the use of pictures in instructional material might be one factor that could affect the development of an ability to integrate the skill of making inferences with the other skills that are performed during reading. The construction of a mental model while reading gives rise to making inferences and depends on the integration of such inferences. This process leads to an enhanced memory of the text and assists comprehension. Four questions were raised: To what extent do 7-9 year old readers draw inferences from pictures rather than text? Is there any difference between male and female readers of this age and in each of these reader groups in how illustrations affect their ability to draw inferences from text? Do pictures hinder 7 -9 year old readers who can decode text well and have a high level of vocabulary knowledge but do not comprehend well (referred to throughout as 'difference poor readers') in their attempts to incorporate general knowledge in their comprehension of text? Do pictures hinder 7-9 year old good readers in the same way? A repeated measures study was designed; 48 children read four stories each, rotated between 4 groups (good readers and poor readers, male and female) so that each story was read with and without pictures. Each child read 2 stories in one condition in the first session and the other 2 stories in the other condition in the second session. The order in which the stories were read was also rotated. The children gave oral answers to questions designed to probe for evidence of inference making after each reading with books open and then re-told the stories with books closed. Results indicated that for these participants, the male difference poor readers were more likely than any in the other groups to make more inferences from the text without the pictures; for some the response also indicated that the nature of their reading experience was profoundly affected. The good readers also tended to make fewer inferences with pictures as did half of the female difference poor readers. The other half seemed to make very few inferences unless pictures were present but these inferences were more influenced by the text than the pictures. These findings suggest that if the aim is to teach comprehension of text then learning materials might be more effective if they did not contain illustrations.