The effects of reader-generated questions on the processing of expository prose

The effects on comprehension, self-monitoring, and retention of science expository prose by at-risk college students using a self-questioning reading strategy were investigated. The subjects were 56 Educational Opportunity Program students enrolled in four sections of a college reading and study skills course. The subjects were predominantly (79%) first-year students, 98% of whom were African-American, Asian, or Hispanic students. Using a quasi-experimental approach, the four sections were randomly assigned to the self-questioning experimental group or the read-reread control group. Subjects were examined on gender, class level, ethnicity, verbal ability, and high school grade point average. The effects of six possible mediating variables--reading ability, logical relationships identification ability, textbook reading strategies use, learning and study strategies use, prior knowledge, and free recall ability--were also examined. The intervention consisted of two weeks of pretesting, three weeks of training/practice with a chemistry chapter, and two weeks of application with a biology chapter. Sessions were conducted twice a week during the last 20 minutes of the class period. Immediate posttesting followed the training and application sessions. Delayed posttesting followed three weeks later. The two 10-item chemistry and biology posttests consisted of text-explicit and text-implicit multiple-choice and short-answer items. Monitoring was measured by comparing self-reported confidence about test performance and actual test performance. Results of the study did not support the three hypotheses that the self-questioning group would exhibit better comprehension by correctly answering more immediate posttest questions, more accurate monitoring by having greater agreement between self-ratings of confidence in test performance and actual test performance, and better retention by correctly answering more delayed posttest questions than the read-reread control group. No significant effects of the proposed mediating variables on posttest performance were found. Examination of questions generated indicated that at-risk students can learn a self-questioning reading strategy. The results suggested that cognitive, metacognitive, and motivational factors all be considered in conducting strategy training and that providing sufficient training and processing time, and using appropriate reading-level materials with model questions is necessary.