Effects of emotional content on working memory capacity

ABSTRACT

Emotional events tend to be remembered better than neutral events, but emotional states and stimuli may also interfere with cognitive processes that underlie memory performance. The current study investigated the effects of emotional content on working memory capacity (WMC), which involves both short term storage and executive attention control. We tested competing hypotheses in a preregistered experiment (N = 297). The emotional enhancement hypothesis predicts that emotional stimuli attract attention and additional processing resources relative to neutral stimuli, thereby making it easier to encode and store emotional information in WMC. The emotional impairment hypothesis, by contrast, predicts that emotional stimuli interfere with attention control and the active maintenance of information in working memory. Participants completed a common measure of WMC (the operation span task; Turner, M. L., & Engle, R. W. [1989]. Is working memory capacity task dependent? Journal of Memory and Language, 28, 127–154) that included either emotional or neutral words. Results revealed that WMC was reduced for emotional words relative to neutral words, consistent with the emotional impairment hypothesis.

KEYWORDS:

• Working memory
• emotion
• OSPAN

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. An additional five participants were outliers on the time spent to complete the OSPAN task (i.e. greater than 3 SD from the mean; OSPAN duration M = 10 m 12 s; SD = 2 m 22 s). We did not specify a priori that we would exclude participants based on this criterion, so they were included in subsequent analyses. Results did not change when these outliers were excluded.

2. In the preregistration document we planned to conduct t tests on the OSPAN scores, but in retrospect the MANOVA seemed more appropriate. Results did not change when WMC scores were analyzed with t tests: for ANL, t (292.87) = 2.36, p = .019, 95% CI [0.50 to 5.47], d = 0.27; for ANU, t (295) = 2.00, p = .047, 95% CI [0.001 to 0.09], d = 0.23; for PCL, t (289.92) = 1.19, p = .233, 95% CI [−0.57 to 2.35], d = 0.14; and for PCU, t (295) = 1.19, p = .236, 95% CI [−0.01 to 0.40], d = 0.14.

3. Word sets included a mixture of both positive and negative words (i.e. no list contained only positive or only negative words), so recall of positive versus negative words was necessarily only part of the set. Therefore, all-or-nothing scoring was not appropriate; we quantified recall for positive versus negative words using partial-credit scoring only.