Better memory for emotional sources? A systematic evaluation of source valence and arousal in source memory

ABSTRACT

Emotion-enhanced memory (EEM) describes the robust memory advantage of emotional over non-emotional stimuli. While extensively investigated with emotional items, it is unclear whether the EEM effect extends to source memory for a neutral item's emotional context. In two pre-registered studies, we systematically manipulated source valence (positive, negative) between participants and source arousal (high, low, neutral-low) within participants. In Experiment 1 (lab study, N = 80), we used emotional sound sources and presented them together with neutral pictures as items. In Experiment 2 (online study, N = 172), we used emotional background pictures with superimposed neutral item words to similarly manipulate source emotionality. Multinomial model-based analysis showed no general effects of valence or arousal on source memory across both experiments. Source memory was impaired for the negative high-arousing source in Experiment 1 but this did not replicate in Experiment 2. Altogether, we conclude that there are no memory-enhancing effects of source emotionality (valence, arousal, or any specific combination thereof) on source memory, dissociating emotionality effects between source and item memory. Additionally, we propose that material-dependent influences carry more weight if the used emotional material is limited in number, as is the case in the standard source-monitoring paradigm employing few sources only.

KEYWORDS:

• Source memory
• emotion-enhanced memory
• valence
• arousal
• multinomial modelling

Acknowledgements

We thank Katja Bitz, Michelle Dörnte, Alina Kias, Franziska Leipold, Paula Schmelzer, and Pia Vogel for help with participant recruitment and data collection.

Disclosure statement

We have no financial or non-financial competing interests that might influence the results reported in this manuscript.

Data availability statement

Data and analysis code (and/or output files) that support the results reported in this manuscript are publicly available in the Open Science Framework repository at https://osf.io/j2axs/?view_only=9f8283891e2e4136921a862e1fcdc982. Both of our experiments were preregistered and can be accessed via https://osf.io/ey64z/?view_only=0772062849314a07badb9046a4bc159d (Experiment 1) and https://osf.io/kupfs/?view_only=fd08a15e7f504b099fb2b70f1c381abd (Experiment 2).

Details on power analysis, material characteristics and exclusion criteria, as well as additional (explorative) time-course analyses are reported in the online supplement.

Additional information

Funding

This work was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) [grant number GRK 2277 “Statistical Modeling in Psychology”].