The Effects of Inducing Emotions on Knowledge Revision Processes

• In Event: Reducing the Impact of Misconceptions

Abstract

Misinformation about vaccines is prevalent and motivates vaccine-hesitancy among new parents (Salmon et al., 2015). The challenge of reducing the impact of such misinformation is compounded by the fact that vaccine misconceptions, and their correction, often have an affective dimension (Leask et al., 2012). With a mixed-factor experimental design, we investigated whether inducing positive or negative emotions during reading would differentially influence previously established knowledge revision processes (Kendeou et al., 2013, 2014).
Undergraduate students (N=64, 65.6% female; Mage=19.0 years, SD=1.04; MGPA=3.48, SD=.45) were recruited to participate for approximately 1-hour. Participants completed prior knowledge and attitude questionnaires, were randomly assigned to either positive, negative, or no emotion induction (neutral) conditions, and self-reported their baseline emotions using the Positive and Negative Affect Scale (PANAS-X; Watson & Clark, 1994); emotions did not differ at baseline, p>.25. After baseline collection, participants in the positive and negative emotion induction conditions watched popular culture videos previously validated to induce emotional experiences. Immediately after viewing, the PANAS-X was completed a second time, which demonstrated treatment fidelity: pairwise t-tests with false detection rate correction showed that positively induced participants reported greater joviality and less negative affect, fatigue, and sadness; conversely, negatively induced participants reported greater sadness and negative affect, and less positive affect, joviality, serenity, and attentiveness. Participants in these conditions then proceeded onto experimental texts. In lieu of videos, neutral condition participants proceeded directly onto experimental texts.
All participants read three Refutation-Explanation texts (RE; texts that explicitly refute misconceptions and provide correct explanations), three Refutation-Non-Explanation texts (RNE; identical structure to refutation texts but without explanations), and three control texts (Non-Refutation-Non-Explanation; NRNE; equivalent content that extended the storyline).
Each text contained a correct outcome sentence (range of 40-43 characters); reading times were recorded for this sentence (Table 1). After reading, participants completed the prior knowledge and attitude questionnaires again. Pairwise t-tests showed that texts overall increased vaccine knowledge and positive attitudes, p’s<.001.
A 3×3 repeated measures ANCOVA showed a significant main effect of Text, F(2, 104)=9.57, p<.001, η2=.155, and a near significant interaction between Text×Induction, F(4, 104)=2.10, p=.086, η2=.075; no significant main effect was found for Induction, p>.20. Planned comparison t-tests showed that RE outcomes overall were read faster than either RNE or NRNE, p’s<.05. Further, negatively induced participants read RE outcomes faster than RNE or NRNE (p’s<.05) and faster compared to RE outcomes read by positively or neutrally induced participants (p’s<.05).
The current study is the first to our knowledge to experimentally examine the effects of emotion on real-time knowledge revision. Planned comparisons showed that negative emotions more than positive or neutral emotions resulted in faster reading times of critical parts of refutation texts, which denotes greater ease of integration and thus successful knowledge revision. We speculate that negative emotions may have augmented processing of correct explanations, which in turn facilitated revision. This interpretation is consistent with a broad literature that demonstrates the stronger impact that negative stimuli have on general processing (Baumeister et al., 2001).

Authors

• Greg Trevors, University of South Carolina

• Catherine M. Bohn-Gettler, College of Saint Benedict

• Bader IMNM Mohsen, University of Minnesota - Twin Cities

• Panayiota Kendeou, University of Minnesota