LaPorte, R., & Voss, J. F. (1974). Paired-associate acquisition as a function of number of initial nontest trials. Journal of Experimental Psychology, 103(1), 117-123.
http://dx.doi.org/10.1037/h0036836
To study the role of testing in paired-associate learning, the testing phase of paired-associate recall trials was replaced by a series of computational tasks. Following 0, 5, 10, 20, or 40 nontest pairing trials, 80 undergraduates received a recall trial and subsequently learned the list to a criterion of 1 errorless trial. Results indicate that although it took an average of 7.5 trials to learn the 12-pair list in the standard condition (Condition 0), mean recall was approximately 10 correct responses following 20 and 40 nontest trials and approximately 7 correct responses following 5 and 10 nontest trials. Mean trials to criterion, not including the nontest trials, was not significantly different among Conditions 0, 5, and 10, although Conditions 20 and 40 yielded significantly fewer trials to criterion than the other 3 conditions. Results support the view that testing facilitates the development of retrieval processes; an additional interpretation is that the test trial provides feedback which influences selective attention and/or rehearsal on the subsequent list presentations, thereby also influencing the acquisition of the associations. (PsycINFO Database Record (c) 2016 APA, all rights reserved)