Perceived memory credibility: The role of details

Abstract

The sharing of personal memories is a unique aspect of the human experience. Humans communicate to provide information, to influence, or even to amuse. How do we distinguish between credible and noncredible narratives? Forensic science has identified race, age, and detail quantity as influential. We do not know how the nature of narrated details impacts believability. We report two studies investigating how detail composition influences credibility ratings using definitions of details suggested by Levine et al. (2002). Internal details are directly connected to the episodic aspects of the memory, while external details refer to semantic facts or depictions not directly related to the main event. A total of 825 participants rated narratives that varied detail number and type for perceived credibility or saliency. Episodic memory details enhanced credibility more than semantic memory details. In addition, within episodic memories, person-related details enhanced credibility more than non-person-related details. Our results suggest a lens through which to view the credibility of what we hear and read.

Keywords: believability; credibility; episodic memory.

Fig. 1.

Ratings for each narrative type. In the Top panel, individual ratings that align with the credibility composite score. From left to right are Accuracy (A), Believable (B), and Confidence in the Truth (C). In the Bottom panel, individual rating scales that align with saliency are Interesting (D) and Pleasant (E). Across all ratings, External Only was rating significantly lower than all other narrative types. Similarly, for credibility ratings, Internal Only and Greater Internal were rated higher than Greater External.

Fig. 2.

Study 1 Internal v External Details. (A) Credibility and saliency composite scores are plotted for all narrative types. Overall, credibility composite scores were rated significantly higher than saliency scores. (B) Post hoc assessments of the relationships between narrative types are conceptually displayed in panel B. For both, Internal Only scores rated higher than External Only scores.

Fig. 3.

Study 1 Composite scores evaluated by detail level. Top panel, credibility composite scores were compared for low, medium, and high detail numbers. Bottom panel, salience composite scores were compared across low, medium, and high detail levels. As can be seen, there is no linear relationship between the number of details and the score rating; however, differences exist across. Bonferroni corrections were applied for post hoc comparisons.

Fig. 4.

Study 2. Internal Detail Stories. For credibility and salience composites, we compared two types of internal narrative types: Non-Person-Related internal details, consisting of time, space, perceptual features of the environment, weather, or external events, and Person-Related details consisting of internal feelings or thoughts, or to other individuals present, and their actions.

Fig. 5.

Study 2. We evaluated how the number and type of internal details influenced ratings for credibility and salience composite scores. NPR refers to Non-Person-Related internal details, and PR refers to Person-Related internal details.

Fig. 6.

Schematic of timeline. All participants read 30 narratives and rated each narrative on 5 scales: accuracy, believability, confidence in the truth, interest, and pleasantness. After, participants completed a demographic screener and questionnaires on health and mood.

Fig. 7.

Study 1 memory narrative types graphically presented. Narratives were composed of internal and external details that differed in detail ratios. In the single detail narratives, memories were composed of only Internal or only External details. The following three types of narrative were composed of differing internal/external detail ratios, included a 1:3 ratio, an approximate 1:2 ratio, and a 2:3 ratio. Greater Internal detail narratives always had more internal than external details, and Greater External detail narratives always had more external than internal details. Our last narrative stimuli had equal quantities of internal and external details intermixed. Participants each read and rated the set of narratives which included at least two representations of all narrative types.