Close replication attempts of the heat priming-hostile perception effect

Abstract

DeWall and Bushman (2009; Experiment 2) reported a study in which participants were exposed to heat-related, cold-related, or neutral (i.e., non-temperature-related) primes prior to reading an ambiguously aggressive vignette. Participants exposed to the heat-related primes judged the vignette's protagonist as more hostile than participants in the cold-priming condition (d=0.67) or neutral-priming condition (d=0.63). This suggests that people mentally associate heat-related constructs with aggression-related constructs. To test the reliability of the effect and to estimate a more precise effect size, the current studies closely replicated DeWall and Bushman in two independent samples, each of which was more than two and a half times greater than the samples in the original study (total N=688). These replication attempts failed to find any evidence that exposure to heat primes affected hostile perceptions relative to the cold primes (ds<−0.06) or neutral primes (ds<0.00). Further, a meta-analysis estimated that the difference in hostile perceptions between those in a heat priming condition and those in a neutral condition was about one-fifth of a standard deviation and not significantly different from zero, d=0.18, 95% CI[−0.09, 0.44]. Thus, I conclude that priming individuals with heat-related constructs does not reliably affect hostile perceptions.