Lewis, Mark2015-08-192015-08-192013-05https://hdl.handle.net/11299/173926University of Minnesota Ph.D. dissertation. May 2013. Major: Educational Psychology. Advisor: Sashank Varma. 1 computer file (PDF); vi, 113 pages.This dissertation proposes a novel approach for improving learning using testing. A growing body of research has demonstrated that testing can be a powerful tool not just for assessing - but also for enhancing learning. However, previous research has only demonstrated that testing can increase past learning (i.e., learning from materials encountered prior to testing). By contrast, this dissertation proposes that testing can also be used to potentiate future learning (i.e., learning from new materials encountered for the first time after testing). According to the proposed model, testing potentiates future learning by producing generative knowledge bases. These generative knowledge bases are composed of featurally-rich memory traces that are fluently retrieved during future text processing, supporting comprehension, learning, and even conceptual change. This proposal is tested in three experiments using naturalistic expository texts. Experiment 1 replicated previous findings of test-enhanced learning and extended them to rich expository texts on topics in psychological science: being tested on a text increased the amount of information participants learned from that text relative to an equivalent amount of restudy, as indicated by delayed tests administered a week later. Experiment 2 demonstrated that testing potentiates future learning and conceptual change: being tested on a text increased the amount of information that participants learned from a new, but related, text that negated, qualified and elaborated the original text. Experiment 3 demonstrated that testing can potentiate future learning as measured by educationally relevant tasks: being tested on a recently read text improved students' critical essays and short answer questions about a new text encountered after testing. Experiment 3 also demonstrated that strongly held beliefs can interfere with the benefits of testing, preventing test-potentiated conceptual change. These results contribute to and extend the emerging view of testing as a valuable tool for promoting learning and represent a new tool for improving learning from expository texts.enTesting EffectText ComprehensionTransferThesis or Dissertation