The effects of computer-administered instructions providing domain or strategy knowledge on the comprehension of familiar and unfamiliar expository text

The purpose of the study was to examine the effects of computer-administered instructions on the comprehension of familiar and unfamiliar college-level material. The instructions addressed two majors issues: (a) the effects of domain-specific knowledge and (b) the effects of strategy knowledge (i.e., knowledge about methods for active, purposeful reading). There were 157 university students who were recruited from psychology classes to participate in the study. There were two instructional conditions and two control conditions for the familiar and unfamiliar domain. The first instructional condition presented background information and information on concepts that were central to the topic. The second instructional condition instructed subjects on when and how to generate questions, summarize, and reread portions of the text. The third condition was a control condition in which subjects only read the text before taking the comprehension tests. The fourth condition was a control condition in which subjects were given the domain instructions and the tests without reading the text. The results were examined in terms of performance on each of three comprehension tests. On the sentence verification test, only a main effect for content familiarity was found. Subjects in the unfamiliar content condition performed better than subjects in the familiar content condition. On the summary test, there was a significant interaction of instruction with content familiarity. Contrary to what was expected, the advantage of the strategy condition over the control was not significant for the familiar content condition. For the unfamiliar content condition both the domain and strategy conditions showed a significant advantage over the control condition, but there was no difference between the domain and strategy. For the inference task, no effects of instructions were found for either condition of content familiarity. The findings provide support for the conclusion that strategy knowledge can be useful for comprehension even in the absence of domain knowledge. The evidence for the efficacy of the domain instructions used in the present study was weak, possibly due to methodological problems. The findings also support the use of multiple measures of comprehension in studies that examine the effects of comprehension instructions