Are There Age-Related Differences in Social Suggestibility to Central and Peripheral Misinformation?

Abstract

Background/Study Context: Dalton and Daneman (2006, Memory, 14, 486–501) showed that young adults can be induced to accept misinformation from a co-witness, even if it contradicts central features of a previously witnessed event. This study investigated whether older adults are also susceptible to social suggestion, and if so, whether to the same or different degree as their younger counterparts. The study also investigated whether participants were more likely to succumb to suggestions delivered by a peer or an older figure.

Methods: Younger and older adults viewed an action video in the presence of a younger or older confederate co-witness. During a postevent discussion, the confederate introduced misinformation about central and peripheral features of the co-witnessed event. Finally, participants responded to true-false statements about the event and rated how confident they were in their decisions.

Results: Older adults were able to correctly reject false statements about an event that had been mentioned during the discussion by the confederate less often than they were able to correctly reject false statements that had not been mentioned, even if the misstatements contradicted central features of the previously witnessed event. However, older adults were no more susceptible to a co-witness's misleading suggestions than were their younger counterparts, and the age of the confederate did not influence the size of the suggestibility effect for younger or older adults.

Conclusion: When baseline memory accuracy (correct rejection rates for unmentioned false information) is controlled, older adults are no more susceptible to misleading suggestions from a co-witness than are their younger counterparts. Age of the confederate did not influence the size of the suggestibility effect and thus provided no support for the predictions that participants are more likely to succumb to misleading suggestions delivered by a peer or by an older authority figure.

Acknowledgments

This research was supported in part by a grant from the Natural Sciences and Engineering Research Council of Canada to M. Daneman. The authors thank Daniel B. Wright and an anonymous reviewer for feedback on an earlier draft.

Notes

¹The two main ways that have been used to introduce postevent misinformation socially is during a pretest discussion (Dalton & Daneman, 2006; Wright et al., 2000, Experiment 2) or during the test phase itself (Wright et al., 2000, Experiment 1). We chose to use Dalton and Daneman's (2006) pretest discussion paradigm because it is more powerful than the sequential test response paradigm in studies that involve small numbers of trials for each condition as ours did.

²For example, in the Linde and Patterson (1964) study, in-group conditions had able-bodied confederates paired with able-bodied participants or disabled confederates paired with disabled participants; out-group conditions had able-bodied confederates paired with disabled participants or disabled confederates paired with able-bodied participants.

Note. Scores are out of 10.

³We used Dalton and Daneman's (2006) procedure for converting a participant's true or false response and confidence rating to a numerical value ranging from 1 to 10 (see also Loftus, 1979b). The suggestibility resistance index (SRI) can be characterized as a function of response accuracy (RA) and confidence rating (CR) as follows: SRI = (CR + 5)^RA × (6 − CR)^(1 − RA). When the response is correct (RA = 1), SRI = CR + 5; in other words, SRI reflects the confidence rating as a bonus added to a base score of 5. When the response is incorrect (RA = 0), SRI = 6 − CR, so that SRI reflects the confidence rating as a penalty subtracted from a base score of 6. For example, if a participant incorrectly responded “true” to a mentioned central false statement (i.e., RA = 0) and gave the response a high confidence rating of 5, then that participant received a suggestibility resistance score of 1 (i.e., SRI = 6 − CR = 6 − 5 = 1), a score that reflected very low resistance to central misinformation and inappropriately high confidence. If a participant incorrectly responded “true” to a mentioned central false statement (i.e., RA = 0) but gave the response a low confidence rating of 2, for example, then that participant received a suggestibility resistance score of 4 (i.e., SRI = 6 − CR = 6 − 2 = 4), a score that reflected moderate resistance to central misinformation and appropriately low confidence.

⁴The same pattern was found for the suggestibility resistance scores (see Figure 2); there was an overall main effect of age of participant, F(1, 76) = 19.81, partial η² = .21, p < .001, but the Participant Age × Mentioned-Unmentioned interaction was not significant, F(1, 76) = 1.98, partial η² = .02, p > .16.

⁵The same pattern was found for the suggestibility resistance scores; there was an overall main effect of age of participant, F(1, 76) = 13.80, partial η² = .15, p < .001, as well as a significant Participant Age × Mentioned-Unmentioned interaction, F(1, 76) = 5.01, partial η² = .06, p < .03.

⁶The fact that participants were at chance in their recognition accuracy for unmentioned peripheral true information (45% for younger participants; 50% for older participants), but were well above chance in their recognition accuracy for unmentioned central true information (89% for younger participants; 88% for older participants), is consistent with the view that participants were much less likely to notice or encode peripheral features of the event than central ones (see Table 2).

⁷The same pattern was found for the suggestibility resistance scores where lower scores reflect lower response accuracy (Table 3); there was a significant Centrality by Mentioned-Unmentioned interaction, F(1, 76) = 53.57, partial η² = .41, p < .001, which showed that suggestibility resistance scores were lower for peripheral true information than for central true information, with the difference being more pronounced when the information was not mentioned by the confederate during the discussion.