Varying retrieval conditions to study survival processing

ABSTRACT

The survival processing advantage is a mnemonic benefit resulting from processing items for their relevance to survival. One explanation of the survival processing advantage is the richness-of-encoding hypothesis: Survival processing enhances retention by generating ideas (elaborative and distinctive processing), increasing the number of retrieval cues. Without retrieval, encoding is futile. Hence, the present experiments varied retrieval conditions – via transfer appropriate processing (TAP) tasks – predicting that the survival processing advantage could be reversed. In Experiment 1a, reducing the transfer appropriateness of survival processing caused significantly lower recognition scores after survival processing than after processing of word associates. Experiment 1b replicated a survival processing advantage and found a survival processing disadvantage. In Experiment 2, survival processing was pitted against a gift desirability task and retrieval mode was varied. Survival processing yielded superior memory on a standard free recall test, but the survival processing advantage was eliminated when an unusual retrieval mode was encouraged. Results affirm the importance of context-dependent retrieval.

KEYWORDS:

• Survival processing
• retrieval conditions
• transfer appropriate processing
• adaptive memory

Acknowledgments

We thank Bradley Dishong, Elizabeth Fitzekam, Destinie Gee, Maren Greve, Austin Guida, Olivia Hankins, Megan Harrison, Madeline Hubly, Sarah Jones, Joseph Meza, Dakota Moore, Courtney Noble, Melanie Shaw, Kayla South-Beck, Celeste Stratton, Brent Tonaki, Clayton Vega, and Emily Winniman for assistance with running participants and/or scoring data.

Data availability statement

We report how we determined our sample size, all data exclusions (if any), all manipulations, and all measures in the study, and we sought to follow JARS (American Psychological Association, 2020). All data, analysis code, and research materials are available in in OSF at https://osf.io/qeh3a/?view_only=eaf4f73f056245cd930b2037a23e8a0a. Data were analyzed using SPSS. This study’s design and its analysis were not pre-registered.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 On the final list, only 5.2% of cues were likely to generate more than one target with a forward strength >0.05. The counterbalancing used should lessen concern over this; nonetheless, see also Footnote 2.

2 We ran a separate analysis on the recognition test data excluding the 5.2% of cues noted in Footnote 1 and found the same results: A 2 conditions (between) by 2 rating/encoding tasks (within) test found a statistically significant interaction, F (1, 91) = 10.48, p = .002, ${\eta }_p^2$ = 0.103. On the standard recognition test, no condition difference was found, t (44) = 1.213, p = .23, 95% CI [−0.032, 0.128], d = 0.181; whereas a sizable condition difference was found on the associates memory test, t (47) = 7.767, p < .001, 95% CI [0.148, 0.251], d = 1.121.

3 Significant results – from accidental data – are reported in this footnote: The wrong power analysis for Experiment 2 was initially conducted (a statistical test of difference between two independent means), leading to 37 extra participants before the mistake was detected. Though data from these extra participants go beyond our stopping rule, we report results of inclusion here because they are thought-provoking: A 2 conditions (between) by 2 encoding tasks (within) test found no statistically significant interaction, F (1, 69) = 2.68, p = .106, ${\eta }_p^2$ = 0.037. A main effect of condition was found, F (1, 69) = 6.07, p = .016, ${\eta }_p^2$ = 0.081; a main effect of encoding tasks was marginally significant, F (1, 69) = 3.71, p = .058, ${\eta }_p^2$ = 0.051. On the free recall test, a clear survival processing advantage was found: The difference in recall between items from the survival encoding task (n = 36, M = 0.356, SD = 0.113) and items from the gift desirability encoding task (n = 36, M = 0.283, SD = 0.114) was statistically significant, t (35) = 2.81, p = .008, 95% CI [0.02, 0.13], d = 0.65. In contrast, on the context (Santa) recall test, the difference in recall between items from the survival encoding task (n = 35, M = 0.275, SD = 0.136) and items from the gift desirability encoding task (n = 35, M = 0.269, SD = 0.114) was not statistically significant, t (34) = .186, p = .853, 95% CI [−0.06, 0.07], d = 0.05.

4 We thank Jim Nairne for bringing this statement to our attention.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.