Gender Threat, Emotional Expressiveness, and Gender Differences in Risk Ratings

Download data and study materials from OSF

Principal investigator:

Branden Johnson

Decision Research

Email: branden@decisionresearch.org

Homepage: http://www.decisionresearch.org/people/johnson/


Sample size: 396

Field period: 2/9/2010-5/28/2010

Abstract

A persistent gap in risk perceptions--men rate risks lower than do women—has remained unexplained. A separate literature suggests one's "manhood" is more vulnerable to questioning than is one's "womanhood"; men under gender norm threats exhibit physical aggression or homophobia. Thus men might rate risks lower because bravery and confidence are masculinity norms; denying high risk reaffirms one's manhood. Threatening gender norms, as in hypothetical test results implying one knows less about stereotypically "male" and "female" topics than most men, might reduce risk ratings among men ) even lower, particularly those for whom being a man—gender identity—is central. Similar manipulation of women would not threaten them and thus not affect women's risk ratings. A test of this hypothesis found the gender gap in risk ratings was greatest (d=0.70) between men who were anxious/aggressive with high gender identity, and women with the same characteristics. However, the manipulation did not work (anxiety/aggression was similar between men in the threat and affirmation conditions), and ANOVA results were not significant. Aside from methodological issues, these results might indicate that risk ratings are not central enough to gender identity to be affected by gender norm threats. An intriguing alternative, however, is that men and women who feel generally vulnerable (as measured by high anxiety/aggression scores unaffected by the manipulation, perhaps from irrelevant or trait sources) and with gender central to their identity will exhibit stereotypical gender norm behavior regardless of specific threats: men will attenuate, and women will amplify, perceived risk.

Hypotheses

Research Question: Why is there a "gender gap" in risk perception, in that men tend to report lower personal and societal risk from technological, natural and social hazards than do women?

Hypothesis: Men give lower risk ratings than women in part due to gender anxiety, so threatening gender identity will reduce men's (but not women's) risk ratings.

Experimental Manipulations

Participants take a cultural-knowledge "test" including items that on their face are more likely to be known by men (e.g., motorcycles) or women (romance novels). This test is followed by a report that their scores indicate that they do or do not have gender-typical knowledge (they scored higher or lower than 72% of same-gender respondents).

Outcomes

Personal risk ratings (nuclear power plants, stress, high-voltage power lines, suntanning, natural disasters, nuclear waste, handguns, food irradiation)

Summary of Results

The risk perception gap was replicated: men on average exhibited lower risk ratings than women in all cases. The experimental manipulation did not work, in the sense that anxiety/aggression scores were similar between men in the threat and affirmation conditions; no difference was found between women in these conditions, as expected. An omnibus ANOVA (Gender * Gender Identity * Anxiety/Aggression) was not significant (p < .17), nor were any other tests (ps > .13). These results on their face indicate that risk ratings are not central enough to gender identity to be affected by gender norm threats.

Methodological issues also could explain this failure in part. The cultural-knowledge test had worked for male undergraduates, but might not have done so for a random sample of U.S. men, and here the manipulation said it represented gender-related knowledge, not gender identity as in the college-based studies. The over-sampling of men for this experiment—reflecting the literature that gender norm threats worked exclusively on men, and particularly those with high gender identity—combined with TESS constraints on overall sample size might have created insufficient power to reveal significant associations.

The gender gap in risk ratings was greatest (d=0.70) between men who were anxious/aggressive with high gender identity, and women with the same characteristics. Thus an intriguing new hypothesis is that men and women who feel generally vulnerable (as measured by high anxiety/aggression scores unaffected by the manipulation, perhaps from irrelevant or trait sources) and with gender central to their identity will exhibit stereotypical gender norm behavior regardless of specific threats: men will attenuate, and women will amplify, perceived risk.