Terrorists brought down the plane!—No, actually it was a technical fault: Processing corrections of emotive information

Abstract

It is well known that people often continue to rely on initial misinformation even if this information is later corrected and even if the correction itself is remembered. This article investigated the impact of emotionality of the material on people's ability to discount corrected misinformation. The focus was on moderate levels of emotionality comparable to those elicited by real-world news reports. Emotionality has frequently been shown to have an impact upon reasoning and memory, but the generality of this influence remains unclear. In three experiments, participants read a report of a fictitious plane crash that was initially associated with either an emotionally laden cause (terrorist attack) or an emotionally more neutral cause (bad weather). This initial attribution was followed by a retraction and presentation of an alternative cause (faulty fuel tank). The scenarios demonstrably affected participants' self-reported feelings. However, all three experiments showed that emotionality does not affect the continued influence of misinformation.

Keywords:

• Continued influence of misinformation
• Memory
• Emotionality

Preparation of this paper was facilitated by a Discovery Grant and an Australian Professorial Fellowship from the Australian Research Council to Stephan Lewandowsky. The lab's website is located at http://www.cogsciwa.com. We thank Klaus Oberauer and Jeff Rouder for comments on an earlier version, E.-J. Wagenmakers and Ruud Wetzels for advice on the Bayesian analysis, and Charles Hanich, Kate Stephanou, Stewart Craig, David Tang, Briony Swire, and Laura Cook for research assistance.

Notes

¹ We use the label “misinformation” for a piece of information that is corrected after its initial dissemination, without meaning to imply that it was initially introduced in order to intentionally mislead the receiver.

² The concept of scepticism refers to a more stable personality trait that facilitates the emergence of specific suspicions and can be measured independently of the misinformation issue under investigation (cf. Lewandowsky et al., 2009).

³ The association of terrorism and Islam has become an increasingly common stereotype in Western countries (Cainkar, 2002; Sheridan, 2006). Thus, choosing an Islamic city of departure was intended to suggest a religiously motivated terrorist attack and hence make the story more emotionally charged (cf. Hayes, Schimel, & Williams, 2008; Solomon, Greenberg, & Pyszczynski, 1991). A database search on factiva.com yielded 14,599 hits searching for “(terrorism OR terrorist) AND bomb AND Tehran” but only 1,393 when replacing Tehran with Auckland (the base rate of hits for “Tehran” and “Auckland” alone was 628,298 and 1,089,924, respectively). The distribution of the baseline-corrected scores differed significantly from an equal distribution, with continuity-corrected χ²(1) = 21,346, p < .001.