The effects of generative testing on text retention and text comprehension

  • Kim Dirkx (Speaker)
  • Liesbeth Kester (Speaker)
  • Kirschner, P. (Speaker)

    Activity: Talk or presentation typesOral presentationAcademic


    Retrieving information is an effective learning strategy for learning facts. In today’s educationalcontext however, more emphasis is on meaningful learning (e.g., comprehension, evaluation). Inthe present experiment, it was investigated if the testing effect can be generalized to moremeaningful testing methods. For that purpose, summarizing was used in this study. Summarizinghas been acknowledged as a useful technique for enhancing comprehension and requires studentsto retrieve information without cues. In the present study, the effects of restudying versussummarizing (as retrieval strategy) on comprehension were investigated. Because the amount ofinitially retrieved information might affect the final test performance, number of words and ideaunits retrieved during the summary condition were also analyzed. The results showed no effectsof summarizing as retrieval strategy compared to restudying on the final test that measuredcomprehension. Thus, in this study no evidence was found for the generalization of a testingeffect to comprehension using summarizing as test. The retrieval effect of a summary test did notlead to enhancement of comprehension. However, significant positive correlations were foundbetween the amount of words and idea units recalled during summarizing and the final test score.The more words and idea units initially recalled the greater long-term performance. Oneexplanation for the results might be that students had difficulty making good summaries and thisinfluenced their learning. Another explanation might be the rather low initial retrieval. Futureresearch might focus on investigating if other test formats are more suited to enhancecomprehension, application or transfer because these educational goals become more and moreimportant in education. An interesting aspect in such research might be the improvement ofinitial retrieval. Additionally, it might be important to include repeated tests or feedback wheninvestigating the testing effect in educational contexts.
    Period30 Aug 2011
    Event title14th Biennial EARLI Conference for Research on Learning and Instruction: Education for a Global Networked Society
    Event typeConference
    Conference number14
    LocationExeter, United Kingdom