Memory for Emotional Scenes Over Time

Introduction

• This study investigates how emotional content affects memory accuracy and susceptibility to misinformation over time.

• Participants viewed positive and negative images, with some exposed to misinformation.

• Memory recall was tested after one week and one month.

Key Concepts

• Misinformation Effect: Exposure to inaccurate information reduces recall accuracy.

• Constructive Memory: Memory is not a perfect replay but is influenced by post-event information.

• Paradoxical Negative Emotion (PNE) Hypothesis: Negative events are well-remembered but more susceptible to misinformation.

Background and Previous Research

• Memory is not a virtual reproduction but is constructive and influenced by post-event information (Loftus, 2005).

• Eyewitness errors have caused injustices (Porter, Campbell, Woodworth, & Birt, 2003; Wilson, 2003).

• False autobiographical events can be implanted (e.g., Hyman & Kleinknecht, 1999; Loftus & Pickrell, 1995; Porter, Birt, Yuille, & Lehman, 2000).

• Loftus' misinformation paradigm (1970s) showed that exposure to misinformation decreases recall accuracy (Weingardt, Loftus & Lindsay, 1995; Loftus, 2005).

• Time affects misremembering (Loftus et al., 1978; Higham, 1998).

• Most eyewitness studies lack emotional content.

Differing Perspectives on Emotional Content and Memory

• Clinical Perspective: Stress impairs memory retrieval (van der Kolk & Fisler, 1995).

• Decay Theory: Stressful memories decay over time due to forgetting and post-event influences (Laney & Loftus, 2005; McNally & Geraerts, 2009).

• Enhanced Memory View: Negative events are better remembered (Porter & Peace, 2007).

Studies on Emotion and Memory

• Charles, Mather, and Carstensen (2003): Negative images were better recognized than positive images by younger adults.

• Porter and Peace (2007): Traumatic experiences were recalled more consistently over 5 years than positive experiences.

• Talarico, Labar, & Rubin (2004): Memory quality is associated with intensity rather than valence.

Emotion and Misinformation Vulnerability

• Negative emotion may increase susceptibility to misinformation.

• Porter, Spencer, and Birt (2003): Negative emotional images were associated with increased susceptibility to major misinformation.

• False memories can be generated for both positive and negative events (Loftus & Pickrell, 1995; Hyman, Husband, & Billings, 1995).

• Porter, Yuille, & Lehman, (1999): The highest rate (56%) of false memories was found for highly negative events.

• Porter, Taylor, and ten Brinke (2008): Participants recalled more true-negative events than true-positive events and recalled false-negative events more than false-positive events.

Paradoxical Negative Emotion (PNE) Hypothesis

• Negative emotion may facilitate memory but heighten susceptibility to misinformation.

• Evolutionary perspective: Negative events should be recalled better to avoid future threats (Porter & Peace, 2007).

• Incorporating information from trusted sources concerning negative events could be adaptive.

• Major details may be more likely incorporated into memory.

• PNE hypothesis predicts negative information will be well remembered but more susceptible to misinformation.

Current Study

• The PNE hypothesis was tested by examining memory for positive and negative images over 1 week or 1 month.

• Some participants were exposed to misinformation.

• The study examined roles of emotionality, post-event misinformation, and time on memory accuracy.

• Predictions: Negative images would be recalled as accurately or more accurately than positive images; negative images would be recalled better after delay; memory for negative images would be more susceptible to misinformation.

Method

• Participants: 80 undergraduate students (66 female, 14 male), mean age = 21.12 (SD = 4.7).

• Random assignment to misinformation (n = 40) or no misinformation (n = 40).

• Follow-up sessions: 1 week (n = 45) or 1 month (n = 35).

• Design: 2 (emotion: positive vs. negative) × 2 (type of information: misleading vs. no misleading) × 2 (latency: 1 week vs. 1 month) X 2 (session: initial vs. follow-up) mixed design.

• Emotion and session were within-subjects variables.

• Type of information and latency were between-subjects variables.

• Materials: International Affective Picture System (IAPS) was used.

• Ten photographs were selected based on emotional valence and arousal.

• Positive images: mean emotional level = 7.24, mean arousal level = 4.93.

• Negative images: mean emotional level = 2.42, mean arousal level = 5.99.

• Images had visible backgrounds and people present.

Misleading and Nonmisleading Questions

• Each image had three sets of questions.

• Misled participants received 10 questions (5 misleading, 5 nonmisleading).

• Major misleading detail: false suggestion of a person, animal, or significant object not present.

• Nonmisled participants received 10 nonmisleading questions.

• All participants answered 15 open-ended questions (Question Set 2).

• The same questions were asked in the follow-up session (Question Set 3).

Procedure

• Participants viewed two images (one positive, one negative) for 30 seconds each.

• A 15-minute questionnaire distracter task was given.

• Participants answered verbally presented questions about each image (Question Set 1).

• After filler questionnaires, participants answered 15 open-ended questions (Question Set 2).

• Follow-up session: participants answered the same 15 questions (Question Set 3).

• Responses coded for accuracy (accurate or inaccurate).

• Accuracy for major misleading details was coded dichotomously.

Results

• A 2 (emotion) × 2 (misinformation exposure) × 2 (latency) × 2 (session) X2 (question type) mixed analysis of variance (ANOVA) was conducted.

• Main effects: Exposure to misinformation, F(1, 74) = 5.68, p < .05, η^2 = .07, and session, F(1, 74) = 24.87, p < .001, η^2 = .25.

• Misled participants were less accurate (M = 45.41%, SD = 17.49) than nonmisled participants (M = 53.84%, SD = 13.09).

• Participants were more accurate initially (M = 52.01%, SD = 16.19) than at follow-up (M = 47.24%, SD = 15.45).

• Interaction: Misinformation exposure × latency, F(1, 74) = 7.48, p < .01, η^2 = .09.

• Misled participants' memory decreased more from 1 week to 1 month.

• Interaction: Exposure to misinformation × question type, F(1, 74) = 10.61, p < .01, η^2 = .13.

• Misled participants were less accurate on questions regarding misleading details.

Incorporation of Major Misleading Detail

• A 2 (misinformation exposure) X2 (emotion) × 2 (session) × 2 (latency) mixed ANOVA was used.

• Main effects: Misinformation exposure, F(1, 75) = 15.01, p < .001, η^2 = .17, and emotion, F(1, 75) = 7.22, p < .001, η^2 = .09.

• Misled individuals were more likely to incorporate major misleading details (M = 0.80, SD = 0.23) than nonmisled participants (M = 0.96, SD = 0.10).

• Major misleading details were more likely incorporated into memory for negative images (M = 0.82, SD = 0.31) than for positive images (M = 0.92, SD = 0.20).

Discussion

• Emotion did not immunize memory from misinformation.

• Misleading questions impaired recall accuracy across emotional conditions.

• Impairment was specific to questions pertaining to misinformation.

• Individuals were more likely to incorporate major misleading details into negative images.

• Negative emotion heightens suggestibility in the presence of major misinformation, consistent with the PNE hypothesis.

• The hypothesis that negative scenes would be more accurately recalled was not supported.

• The passage of time impaired recollection.

• Memories of misinformed participants were particularly fragile.

• Misinformation may accelerate the natural decay of memory over time.

Conclusion

• Predictions concerning the impact of emotion on memory accuracy were partially supported.

• Negative emotion heightens suggestibility in the presence of major misinformation relative to positive emotion, consistent with the PNE hypothesis.

• Improper questioning techniques should be avoided to reduce erroneous testimony.