The effects of adjunct questions on learning from text inconsistent with prior knowledge

Prior knowledge can interfere with children's learning if it is inconsistent with the to-be-learned information. Adjunct questions have been effective in increasing retention, particularly for directly questioned information. The present study was conducted to investigate the effect of adjunct questions directed toward textual material likely to be inconsistent with children's background knowledge. It was hypothesized that children who read the text accompanied by factual questions would learn the presented information better than children who read the passages without accompanying questions. It was also hypothesized that higher order adjunct questions would be superior to factual questions because the higher order questions would result in deeper processing and enable more thorough integration of new information with existing knowledge.; To test these hypotheses, fifth graders (N = 320) read passages containing information likely to be inconsistent with their prior knowledge and responded to prequestions, postquestions, both pre- and postquestions, or no questions. The questions were factual or higher order.; Children who responded to factual questions scored no better on a multiple-choice posttest than those receiving no questions. Children in all three higher order question groups scored below children who received factual pre- and postquestions. Subjects earned lower scores on items related to the misconceptions than on those unrelated to the misconception. Above-average readers performed better than average readers, who scored higher than below-average readers. Reading achievement did not interact with question position or level. Performance of a passage-dependency comparison group that took the posttest without reading passages confirmed that children needed to read the material to be able to respond above chance levels. The results suggest that factual adjunct questions do not improve learning for textual material that is inconsistent with prior knowledge. Findings also show that higher order questions can interfere with learning. Processing explanations to account for the findings are discussed.