The effect of mood on confidence in false memories

Abstract

The goal of this study was to examine the effect of mood on suggestibility in the misinformation paradigm. To investigate the relative effects of valence and arousal, as well as affect-specific influences, six mood conditions were included: positive mood with low/high arousal (serene/happy), negative mood with low/high arousal (sad/angry), neutral mood, and a control condition. Participants watched a movie and were exposed to misleading information by means of a narrative. Memory was tested in a surprise forced-choice recognition task, with confidence judgements. The mood induction procedure was shown to be effective. A significant misinformation effect confirmed that participants were misled by the false information provided. Mood did not affect susceptibility to the misinformation effect, but did significantly influence participants’ belief in their false memories. Feeling sad induced the highest confidence ratings. Results are discussed in terms of different problem-solving strategies associated with discrete affective states, and have implications for both legal and clinical settings.

Keywords:

• Arousal
• Confidence
• False memory
• Misinformation paradigm
• Mood
• Valence

Acknowledgments

The research reported in this paper was supported by the Research Foundation–Flanders (FWO), of which IVD is a Postdoctoral Fellow. The authors would like to thank Melanie Takarangi and Maryanne Garry for providing the misinformation materials, and Yves Corson for providing the materials for the mood induction.

Notes

¹“Mood” is typically defined as a diffuse affective state, whereas an “emotion” is more short-lived and reflects an underlying judgement of a particular object or situation (e.g., Forgas, 1995).

²Two in the serene condition, one in the happy condition, two in the sad condition, one in the angry condition, one in control condition 1, and one in control condition 2. All of these participants already showed extreme mood scores at the start of the experiment.

³One in the serene condition, one in control condition 1, and one in control condition 2.

⁴A 6 (mood)× 2 (misinformation)× 2 (gender) analysis of variance additionally revealed a main effect of gender, F(1, 277) = 7.92, MSE=0.06, p=.005, with lower accuracy for female than for male participants. Importantly, however, there were no significant interactions. There were also no main or interaction effects of gender on confidence in either correct or false memories.

⁵An analysis of variance on conditionalised averages for macrosubjects (procedure cf. Loftus et al., 1989) revealed similar results. Macrosubjects were created by combining data from three or four subjects in order to eliminate empty cells and increase stability of measurement.