The impact of event vividness, event severity, and prior paranormal belief on attributions towards a depicted remarkable coincidence experience: Two studies examining the misattribution hypothesis.

Two studies examine the impact event vividness, event severity, and prior paranormal belief has on causal attributions for a depicted remarkable coincidence experience. In Study 1, respondents (n = 179) read a hypothetical vignette in which a fictional character accurately predicts a plane crash 1 day before it occurs. The crash was described in either vivid or pallid terms with the final outcome being either severe (fatal) or non-severe (non-fatal). Respondents completed 29 causal attribution items, one attribution confidence item, nine scenario perception items, a popular paranormal belief scale, and a standard demographics questionnaire. Principal axis factoring reduced the 29 attribution items to four attribution factors which were then subjected to a 2 (event vividness) × 2 (event severity) × 2 (paranormal belief) MANCOVA controlling for respondent gender. As expected, paranormal believers attributed the accurate crash prediction less to coincidence and more to both paranormal and transcendental knowing than did paranormal sceptics. Furthermore, paranormal (psychokinesis) believers deemed the prediction more reflective of paranormal knowing to both (1) a vivid/non-fatal and (2) a pallid/fatal crash depiction. Vividness, severity, and paranormal belief types had no impact on attribution confidence. In Study 2, respondents (also n = 179) generated data that were a moderately good fit to the previous factor structure and replicated several differences across attributional pairings albeit for paranormal non-believers only. Corresponding effects for event severity and paranormal belief were not replicated. Findings are discussed in terms of their support for the paranormal misattribution hypothesis and the impact of availability biases in the form of both vividness and severity effects. Methodological issues and future research ideas are also discussed.