# Possessing potential weapons (still) heightens anger perception: Replicating and extending a test of error management theory

## TL;DR

This study replicates findings that people perceive individuals holding potentially lethal objects as angrier and more dangerous, but does not find that this perception extends to their friends.

## Contribution

The study replicates prior EMT findings and explores a novel 'reverse-halo' effect without finding support for it.

## Key findings

- Individuals holding potentially lethal objects are perceived as angrier and more anger-prone.
- Emotion ratings of friends were significantly correlated, suggesting homophilic similarity.
- No evidence was found for a 'reverse-halo' effect extending anger perception to friends.

## Abstract

Error Management Theory (EMT) hypothesizes that humans are functionally biased to err on the side of the least costly mistake when making judgments under uncertainty. Applying EMT to emotion perception, previous studies found that people perceive individuals holding everyday objects with potentially lethal affordances to be both higher in state anger and more anger-prone, relative to individuals holding control objects. Here, we conduct a direct replication of one such study examining these effects [13] and exploratorily test whether friends of an individual possessing a potentially lethal tool are also perceived as angrier—and hence more dangerous—in a “reverse-halo” effect. Participants (N = 476) were presented an image of an individual depicted as an avid cook and either holding or near a conventional kitchen knife, then asked to rate their emotional states and trait dispositions. Participants were next presented with an image of a second individual framed as a close friend of the cooking enthusiast and asked to assess their emotional states and dispositions. The results replicated prior findings that individuals depicted holding everyday objects with incidentally lethal affordances are perceived as angrier, more prone to anger, and more socially unpleasant. However, we did not find evidence for a “reverse-halo” effect extending to friends. Instead, all of the state and trait emotion ratings of the two individuals were significantly correlated, consistent with an inference (orthogonal to error management motivations) of homophilic similarity in the affective profiles of friends. These results are discussed as they inform prior EMT research and motivate further study of the determinants of emotion perception.