Effects of Spacing of Item Repetitions in Continuous Recognition Memory: Does Item Retrieval Difficulty Promote Item Retention in Older Adults?

Abstract

Background/Study Context: Older adults exhibit an age-related deficit in item memory as a function of the length of the retention interval, but older adults and young adults usually show roughly equivalent benefits due to the spacing of item repetitions in continuous memory tasks. The current experiment investigates the seemingly paradoxical effects of retention interval and spacing in young and older adults using a continuous recognition memory procedure.

Methods: Fifty young adults and 52 older adults gave memory confidence ratings to words that were presented once (P1), twice (P2), or three times (P3), and the effects of the lag length and retention interval were assessed at P2 and at P3, respectively.

Results: Response times at P2 were disproportionately longer for older adults than for younger adults as a function of the number of items occurring between P1 and P2, suggestive of age-related loss in item memory. Ratings of confidence in memory responses revealed that older adults remembered fewer items at P2 with a high degree of certainty. Confidence ratings given at P3 suggested that young and older adults derived equivalent benefits from the spacing between P1 and P2.

Conclusion: Findings of this study support theoretical accounts that suggest that recursive reminding and/or item retrieval difficulty promote item retention in older adults.

Acknowledgments

This work was supported by grants AG11451 and AG034464 from NIA and grant MH069938 from NIMH. The authors thank Kim Talbott for assistance with data collection, Peter G. Kelley for assistance with the programming algorithms for stimulus presentation, and Matthew Prull for comments on an earlier draft.

Aslı Kılıç is now at the Department of Psychology, Koç University, Istanbul, Turkey. Marc W. Howard is now at the Department of Psychology, Boston University, Boston, Massachusetts, USA.

Notes

Note. Data are arrayed by lag and retention interval ranges for comparisons with Tables 2 and 3; data should not differ by lag and retention interval ranges because these variables refer to the effects of the numbers of items occurring between P1 and P2, and between P2 and P3, respectively. Entries indicate false-alarm rates because all of the words are “new” at P1 and P(9) is the highest confidence “old” response. Numbers in parentheses are standard errors of the mean.

Note. Numbers in parentheses are the standard errors of the means. Lag refers the numbers of items that were presented between P1 and P2. Retention interval refers to the numbers of items that were presented between P2 and P3, and thus any influence on the data in this Table is uncontrolled. Probability of highest confidence responses decreases as a function of lag length, especially for older adults (i.e., entries in bold).

Note. Entries in parentheses are the standard errors of the mean. For young and older adults, the probability of highest confidence responses increased from lag ranges of 1–3 to 6–10 (i.e., a spacing effect, entries in bold).

¹The absence of any reliable effects of lag and retention interval at P1 can be taken to suggest the absence of changes in response bias across the test session.

²The same pattern of results was obtained in an analysis that excluded the primacy buffer in each block. The primary buffer consisted of all of the occurrences of the first 10 items on each list.

³A similar pattern of qualitative results was obtained in an analysis that excluded participants who had a probability higher than .9 of responding 9 to the second presentation of the words.