Moral disengagement and self-reported harassment proclivity in men: the mediating effects of moral judgment and emotions

ABSTRACT

Three online studies investigated the association between moral disengagement and men’s self-reported harassment proclivity. Participants (total N = 336) were required to read a vignette depicting either quid pro quo harassment (studies 1 and 2) or hostile work environment harassment (study 3). A salience manipulation was used in each study to explore the causal directionality of this association. The mediating effects of moral judgment, negative affect (guilt and shame) and positive affect (happiness) about the harassment were also assessed as participants were asked to imagine themselves as the harassment perpetrator. Across the three studies, it was shown that moral disengagement had an indirect effect in predicting men’s proclivity to harass by lowering their moral judgment and negative affect about the harassment, conversely amplifying positive affect. Overall, the findings support social cognitive theory, indicating that moral disengagement may enable people to self-regulate their own behavioural inclinations to harass.

KEYWORDS:

• Moral disengagement
• sexual harassment
• sexual harassment proclivity
• workplace
• social cognition
• emotions

Acknowledgements

We are grateful to the anonymous reviewers for their helpful comments and suggestions on an earlier version of this article.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. It is important to acknowledge that men also experience sexual harassment at work (see Berdahl, 2007b; Holland et al., 2015; Stockdale et al., 2004) either from female or same-sex perpetrators. The current studies are situated only in the context of male-perpetrated sexual harassment of women. This is because it is statistically the most frequent perpetrator-victim constellation (see McDonald, 2012; O’Leary-Kelly, Bowes-Sperry, Bates, & Lean, 2009). However, in the general discussion section, we mention the need to examine other perpetrator-victim dyads in future research.

2. Rape myths have been defined as “descriptive or prescriptive beliefs about rape (i.e., about its causes, context, consequences, perpetrators, victims, and their interaction) that serve to deny, downplay or justify sexual violence that men commit against women” (Bohner, 1998, p. 14).

3. Prolific Academic is a crowdsourcing platform that is used to conduct psychological research online. It is increasingly used as a suitable alternative to Amazon Mechanical Turk (see Peer, Brandimarte, Samat, & Acquisti, 2017) and enables registered users to participate in studies in return for monetary reward. Crowdsourcing platforms such as Prolific Academic and Amazon MTurk are frequently used by social scientists to recruit and compensate participants (see Goodman, Cryder, & Cheema, 2013). These services offer various advantages to researchers such as providing access to samples that are demographically diverse and have more relevant work experience (Behrend, Sharek, Meade, & Wiebe, 2011). Researchers have also demonstrated that crowdsourcing platforms produce highly reliable and valid data that are equivalent or better in quality to data collected using traditional Internet participant pools and university student samples (see Behrend et al., 2011; Berinsky, Huber, & Lenz, 2012).

4. The measures of moral disengagement and moral judgment displayed moderate negative correlations in each of the three studies (study 1: r = −.54, p < .001; study 2: r = −.52, p < .001; study 3: r = −.38, p < .001). Exploratory factor analyses were performed on the combined items of both scales which consistently produced a clear two-factor solution. The eight items of the moral disengagement scale loaded strongly on to factor 1, whereas the thirteen items of the moral judgment scale loaded strongly on to factor 2 revealing independent clusters. These analyses indicate that our measures of moral disengagement and moral judgment are empirically separable as distinct constructs. Full details of these factor analyses can be obtained from the first author.

5. Bootstrapping is a nonparametric resampling procedure. It is widely considered to produce more accurate estimates of indirect effects because it does not impose the assumption of normality of the sampling distribution. Bootstrapping involves repeated resampling of the available data to create an empirical approximation of the sampling distribution. Estimates of indirect effects are calculated with greater precision by constructing confidence intervals that are corrected for bias and accelerated. In each of our analyses, an indirect effect was considered significant when the bootstrapped confidence interval did not contain zero (Hayes, 2009; Preacher & Hayes, 2008).