Exploring the forgetting mechanisms in working memory: Evidence from a reasoning span test

Abstract

In working memory (WM) span tests participants have to maintain to-be-remembered information while processing other, potentially distracting, information. Previous studies have shown that WM span scores are greater when span lists start with a long processing task and end with a short processing task than when these processing tasks are presented in the reverse order (e.g., Towse, Hitch, & Hutton, 2000). In Experiment 1, we obtained a similar stimulus order effect in a reasoning span test, using reasoning sentences that were equated for length in terms of the number of constituent words, but which differed in processing complexity; span scores were greater when lists began with a complex sentence and ended with a simple sentence than when this stimulus order was reversed. In Experiment 2, the stimulus order effect was not found when processing duration was held constant while sentence complexity was varied using a computer-paced moving window presentation paradigm. These results suggest that duration-based constraints can affect degree of forgetting independently of the load generated during processing phases in WM span performance and therefore imply that time-related forgetting can occur in WM span tests, particularly when the difficulty of the processing component blocks active maintenance of to-be-remembered material.

Keywords:

• Working memory
• Reading span
• Task switching
• Resource sharing
• Short-term forgetting
• Stimulus order effect

Acknowledgments

Satoru Saito is a Visiting Fellow at the Department of Experimental Psychology at the University of Bristol.This research was supported by a Grant-in-Aid for Scientific Research (Project 14710082) from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) in Japan, by the 21st Century Center of Excellence Program (D-10 to Kyoto University), MEXT, Japan, by Canon Foundation in Europe to Satoru Saito, and by a cooperative group component grant from the United Kingdom Medical Research Council to Chris Jarrold and Alan Baddeley (Grant G0000258, within Cooperative Group Grant G9901359). We thank Nelson Cowan for helpful comments on an earlier version of this paper.

Notes

¹ In Experiment 1, we obtained significant effects of the affirmative–negative contrast and of the active–passive contrast, perfectly replicating the results of Baddeley and Hitch (1974). The time difference of the effect was larger between the affirmative and the negative sentences (1,867 ms) than between the active and the passive sentences (741 ms), supporting the validity of the selection of simple and complex problems in the current study.

² Although there are concerns as to whether temporal distinctiveness or other temporal factors can explain data from immediate serial recall (see Lewandowsky, Duncan, & Brown, 2004), these concerns do not preclude the possibility that temporal distinctiveness accounts for phenomena in longer term situations like the continuous distractor paradigm (see Lewandowsky, Brown, Wright, & Nimmo, 2006). It might be worth noting that the duration manipulation in our study cannot separate the temporal distinctiveness explanation from a temporal decay account because we manipulated only retention durations and not interitem intervals. However, we incline towards the temporal distinctiveness account to explain the stimulus order effect reported here for two reasons. First the phenomenon seems to occur in a long-term situation, where other evidence suggests that interference rather than temporal decay explanations are better placed to account for forgetting, and, second, a temporal distinctiveness account might provide a better explanation of other related phenomenon (e.g., Unsworth & Engle, 2006a, 2006b).

Additional information

Notes on contributors

Satoru Saito

Satoru Saito is a Visiting Fellow at the Department of Experimental Psychology at the University of Bristol.