Date of Award

Summer 8-15-2013

Author's School

Graduate School of Arts and Sciences

Author's Department


Degree Name

Doctor of Philosophy (PhD)

Degree Type



Retrieval practice directly enhances later memory of tested material, a robust effect known as the testing effect (Roediger & Karpicke, 2006b). Numerous experiments have provided support for this effect. However, another important effect of retrieval practice has received far less attention. Retrieval practice can also indirectly enhance learning by potentiating subsequent encoding of tested material, an effect known as test-potentiated learning (Izawa, 1966). Although introduced over four decades ago, little is known about how and when tests enhance subsequent encoding, information that has both practical and theoretical importance. The aim of this dissertation was to enhance understanding of test-potentiated learning by answering two fundamental questions: First, which aspect(s) of prior tests (unsuccessful retrieval, successful retrieval, and/or spacing) cause subsequent enhanced encoding (Experiment 1) and second, can items that are correctly retrieved prior to restudying benefit from test-potentiated learning (Experiment 2)?

Experiment 1 answered the first question by varying the amount of unsuccessful and successful retrieval subjects engaged in prior to restudying the material. Further, the lag between study periods, task engagement, and exposure to recallable items were held constant across conditions to equate the effect of spacing. Repeated unsuccessful retrieval enhanced the benefit of the restudy trial for initially incorrect items. The implications of this finding for the theoretical processe(s) driving test-potentiated learning are discussed.

Experiment 2 tested the hypothesis that items correctly retrieved prior to restudying can also benefit from test-potentiated learning as evidenced by later enhanced recall. A restudy opportunity has been shown to enhance later recall of previously retrieved items, especially items retrieved with low confidence (Butler, Karpicke, & Roediger, 2008). That is, restudying enhances the effect of prior retrieval. Prior retrieval may also enhance the effect of subsequent restudy. This hypothesis was tested by varying the number of prior tests and whether or not subjects had a restudy opportunity. Further, confidence ratings were collected. Restudying was found to benefit low-confident correct items, but the number of prior tests did not modify this effect suggesting that correct retrieval prior to restudying does not enhance the effect of restudying on later recall. Together, these experiments indicate that unsuccessful retrieval attempts made during prior testing are the driving force behind test-potentiated learning.


English (en)

Chair and Committee

Kathleen B. McDermott

Committee Members

David A. Balota, Joe Barcroft, Susan M. Fitzpatrick, Mark A. McDaniel, Henry L. Roediger, III


Permanent URL:

Included in

Psychology Commons