The role of attention in retrieval practice

Joshua W Whiffen, Purdue University


Dividing attention during encoding is detrimental to learning. In contrast, dividing attention during retrieval appears to have very little effect on recall. However, very few studies have investigated whether dividing attention during initial recall has an impact on subsequent retrieval attempts. Research on retrieval practice has clearly shown that retrieval is an active process that leads to important changes in memory that ultimately enhance long term retention. However, it has yet to be established exactly how retrieval practice derives its benefits. One possibility is that retrieval involves the reinstatement of temporal context, which leads to the updating or encoding of additional context features that can ultimately be used during a later retrieval attempt to help restrict the search set and retrieve the target memory. If this perspective is correct then certain predictions can be made about the influence that dividing attention may have on retrieval practice. Specifically, dividing attention is known to disrupt encoding process, thus dividing attention may disrupt updating of additional context features and therefore prevent the encoding of additional context features. This should then reduce the mnemonic vii benefit of retrieval. Further, this reduced mnemonic benefit should be accompanied by lower scores on measures of temporal context on the criterial test. The following experiments sought to test these predictions. In Experiment 1 subjects either restudied or practiced retrieval while under divided or full attention. Performance on a delayed free recall test revealed a retrieval practice effect, but no effect of dividing attention. Also, there were initial differences in recall between the retrieval groups. However, there were also differences on the measures of temporal context. Overall, due to the difficulty of drawing conclusions about final test performance when there were differences on an initial test, some key changes were made to the next experiment. Experiment 2 used a similar design to Experiment 1 but with some changes to try and avoid any differences in initial performance. There were still initial differences between retrieval groups, although they were reduced thus suggesting that dividing attention was in fact disrupting retrieval. At the same time, there were differences on the final test, such that retrieval groups outperformed study groups, and there was also an effect of attention. After accounting for differences in initial test performance there appeared to be no effect of dividing attention on the mnemonic benefit of retrieval. Also, this experiment again revealed differences in temporal context that was in line with what was predicted by the episodic context account. Experiment 3 sought to again try and reduce initial differences in performance but to also provide a strong test of the episodic context account. The procedure was generally the same as the previous experiments except instead of a final free recall test there was a final list discrimination test. The results showed that there were no differences in initial recall, but also that there were no overall differences in list discrimination performance between any of the groups. Overall, some support for the predictions made by the episodic context account was found in this set of experiments although many questions remain.




Karpicke, Purdue University.

Subject Area

Behavioral psychology|Experimental psychology|Cognitive psychology

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server