The following commentary was inadvertently omitted from Volume 12, Issue 1, as a response to the focal article https://doi.org/10.1017/iop.2018.155, “#Ustoo: How I-O Psychologists Can Extend the Conversation on Sexual Harassment and Sexual Assault Through Workplace Training.” SIOP regrets the error.

We completely agree that the issue raised in the focal article by Medeiros and Griffith (Reference Medeiros and Griffith2019) is extremely important and timely, and we commend the authors for spearheading this discussion. Although we agree with the spirit of this focal article, we wish to provide an important clarification to the perspective offered by the authors. Medeiros and Griffith focus their discussion on sexual harassment that is limited to sexual behaviors enacted toward victims (e.g., sexual advances or sexual assault). Specifically, they drew heavily on the literature on college campus sexual assault, mentioned that sexual harassment and sexual assault exist on a continuum and are characterized by an “escalating nature,” and referenced numerous explicit sexual behaviors in their examples (e.g., sexual advances, sexual assault). Their discussion of training interventions was also focused completely on preventing sexually motivated infractions.

Although workplace sexual harassment may involve sexualized comments or behavior, contemporary frameworks conceptualize sexual harassment as being not about sex but about maintaining power and social status on the part of the harasser (Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011). In the following sections, we discuss this perspective in detail, highlighting three key implications of including a broader “sexual-harassment-as-sexism” definition and discussing how this perspective may change some of the recommendations in the focal article.

Sexual harassment may not look like or be motivated by sexual intent

Sexual harassment was traditionally defined as involving unwanted sexual behavior committed mostly by men directed mostly toward women. It was thought that such behavior was caused by biological factors—specifically men’s purportedly stronger sex drive and desire for sexual expression and gratification (Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011; Tangri, Burt, & Johnson, Reference Tangri, Burt and Johnson1982). It later became clear, however, that sexual harassment takes more sexist than sexual forms, and scholars and U.S. law now define sexual harassment more broadly as behavior that “derogates, demeans, or humiliates an individual” or “creates an intimidating, hostile, or offensive work environment” based on that individual’s sex (Berdahl, Reference Berdahl2007a, p. 644; Equal Employment Opportunity Commission [EEOC], 1980). It is important to note that nonsexual behaviors such as sabotage, ridicule, intimidation, ostracization, and sexist jokes or comments are included in this definition and that this broader, gender-role–based definition is consistent with others (including court rulings) that do not limit sexual harassment to sexual behaviors (cf., Bildt, Reference Bildt2005; Franke, Reference Franke1997; Oncale v. Sundowner Offshore Services, Inc., 1998; Schultz, Reference Schultz1998, as cited in Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011).

The first key implication of this definition is that the most common forms of sexual harassment may not be sexual at all. Indeed, the most common form of sexual harassment involves sexist comments, jokes, and materials that “alienate and demean victims based on sex rather than solicit sexual relations with them” (see Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011). Sexual coercion is actually the rarest form of sexual harassment, with most surveys finding that at most 6% of respondents report quid pro quo harassment (being bribed or threatened to engage in sexual behavior at work; Fitzgerald et al., Reference Fitzgerald, Shullman, Bailey, Richards, Swecker, Gold and Weitzman1988; Raver & Gelfand, Reference Raver and Gelfand2005). In comparison, the same and other studies found that between 35% and 42% of respondents experienced gender-based harassment; only 1% experienced sexual coercion and 15% experienced sexual attention (Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011; Fitzgerald et al., Reference Fitzgerald, Shullman, Bailey, Richards, Swecker, Gold and Weitzman1988; Spann, Reference Spann1990). This can contribute to a situation in which people do not always acknowledge or appraise certain sexist behaviors as constituting “sexual harassment” despite the fact that such behaviors qualify as sexual harassment under contemporary definitions and can result in negative outcomes (Ilies, Hauserman, Schwochau, & Stibal, Reference Ilies, Hauserman, Schwochau and Stibal2003).

Second and relatedly, modern perspectives of sexual harassment hold that harassers are motivated by a desire to maintain gender inequality (Berdahl, Reference Berdahl2007b). In other words, even sexual harassment that appears to be completely sexual in nature (and not gender harassment; i.e., sexual coercion) is a form of hostility triggered by those in higher status positions (typically men, but not always) to protect or enhance their identities in those higher status roles when those identities are threatened (Berdahl, Reference Berdahl2007a; Maass, Cadinu, Guarnieri, & Grasselli, Reference Maass, Cadinu, Guarnieri and Grasselli2003). By highlighting the sexual aspects of the targets, harassers simultaneously delegitimize task-relevant characteristics such as competence, intelligence, or expertise. Indeed, Berdahl (Reference Berdahl2007a) found that such hostile work environments are motivated by hostility toward people (mostly women) who violate gender ideals rather than by desire for those that meet them (see also Dall’Ara & Maass, Reference Dall’Ara and Maass1999; Franke, Reference Franke1997; Maass et al., Reference Maass, Cadinu, Guarnieri and Grasselli2003). Rather than from sexual desire, then, sexual harassment most commonly stems from a desire to maintain social status or power. This places the cause of sexual harassment in the realm of sociological factors such as traditional gender hierarchies and individual psychological motivations such as retribution (e.g., a desire to punish a gender-role violator) or self-presentation (e.g., to prove one’s masculinity).

With respect to interventions, these two implications suggest that the content of trainings designed to reduce sexual harassment should also include attempts to reduce gender-based antipathy in the work context. The authors of the focal article rightly discuss the importance of beginning with a needs analysis. We recommend a needs analysis that includes the antecedents of sexual harassment as sexism as well as employee reports of previous sexual harassment experiences. The Sexual Experiences Questionnaire (SEQ) is commonly used to assess sexual harassment in organizations (Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011; Fitzgerald, Gelfand, & Drasgow, Reference Fitzgerald, Gelfand and Drasgow1995; Fitzgerald et al., 1997) and has been adapted and modified for use in many different contexts (Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011). We recommend that pretraining assessments use items of the SEQ that are related to experiences of gender harassment. Following the recommendations of Berdahl and Raver (Reference Berdahl, Raver and Zedeck2011), we also suggest that subjective appraisals (how an experience made one feel) and outcomes (what well-being or performance outcomes have been experienced) be included in the pretraining assessment. A successful needs analysis will use measurements that allow trainers to interpret what the trainees are experiencing as well as the job context and organizational environment in which they are operating.

When designing the intervention (following a needs assessment), there are several key considerations. First, gender harassment should be considered when deciding content relevance. Training content should help trainees identify gender harassment by teaching that sexual harassment is not entirely motivated by sexual desires but instead can be motivated by a desire to maintain social status and power and enforce gender norms (Berdahl, Reference Berdahl2007a). Training content should also focus on the different antecedents of gender harassment, reporting gender harassment, and bystander intervention. Moreover, training focused on gender harassment should promote empathy toward the diverse victims of gender harassment. Training transfer has been found to be more successful when the training focuses on empathy (Roehling & Huang, Reference Roehling and Huang2018). Therefore, there will be a higher likelihood of training transfer from gender harassment training if participants are aware that victimhood is not limited to women. Successful training design is dependent on having content related to gender harassment.

Second, the practice and feedback portion of the intervention will be impacted by adopting the sexual-harassment-as-sexism framework. The framework proposes that providing trainees with realistic scenarios and prompting them to personally respond will increase the likelihood of training transfer. Acknowledging this assumption, we propose that role-playing scenarios should incorporate the nontraditional views of sexual-harassment-as-sexism. Similarly, sexual-harassment-as-sexism should affect behavioral-based modeling and error-based examples. The authors suggest that behavior-based modeling and error-based examples can be implemented by providing trainees with behavioral examples to observe and assess. We suggest that behavioral examples should be accurate representations of sexual-harassment-as-sexism, and questions should provoke discussions about sexual-harassment-as-sexism. This part of the framework is crucial to training transfer, as sexual harassment training may initiate attitude change if interactive and experiential training methods are used (Perry, Kulik, & Field, Reference Perry, Kulik and Field2009).

It is also worth noting that measurement will change with the nature of the training. Post-training measures and evaluations should focus on measuring behaviors consistent with gender harassment. For an accurate description of training effectiveness, measurement needs to be consistent with the pretraining analysis and the content of the training. Training effectiveness and training transfer can be measured using accurate measurements of gender harassment behaviors.

Sexual harassment varies by gender and contextual influences

Third, although most victims of sexual harassment are women and most perpetrators are men, targets and perpetrators can be of any gender, and harassment between them can be directed toward the same or opposite gender. Sexual harassment was originally thought to be perpetrated primarily by supervisors against subordinates but is actually most often perpetrated by coworkers (e.g., Farley, Reference Farley1978), customers (e.g., Barling, Rogers, & Kelloway, Reference Barling, Rogers and Kelloway2001), and subordinates (e.g., DeSouza & Fansler, Reference DeSouza and Fansler2003). Sexual harassment filings by men to the EEOC increased 9% between 1992 and 2008 (Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011), and when men are harassed, they are as much if not more often harassed by men than as by women (Berdahl, Reference Berdahl2003; Magley, Waldo, Drasgow, & Fitzgerald, Reference Magley, Waldo, Drasgow and Fitzgerald1999; Waldo, Berdahl, & Fitzgerald, Reference Waldo, Berdahl and Fitzgerald1998). This perspective helps explain nonsexual same-sex harassment in which men punish other men for not being “man enough” (Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011) or situations in which men harass women in male-dominated industries or professions or who hold egalitarian views rather than traditional gender role attitudes (Berdahl, Reference Berdahl2007b; Dall’Ara & Mass, Reference Dall’Ara and Maass1999; Maass et al., Reference Maass, Cadinu, Guarnieri and Grasselli2003; as cited in Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011). Additionally, although harassment by coworkers is most common across most industries, it varies; a study found that in service-oriented industries, sex-based harassment by customers has been cited as the most common (Barling et al., Reference Barling, Rogers and Kelloway2001; Berdahl, Reference Berdahl2003; Gettman & Gelfand, Reference Gettman and Gelfand2007). Further consideration of organizational context highlights that occupational sex composition (the extent to which an occupation is typically held by women or by men) and work group sex ratio (the proportion of men and women in the immediate work context)—together known as job-gender context—have been identified as predictors of sexual harassment (Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011). Male-dominated environments are often associated with “locker room” cultures that promote sexual harassment behaviors (e.g., Ely & Meyerson, Reference Ely and Meyerson2008; Harrell & Miller, Reference Harrell and Miller1997), which are disguised as “everyday behaviors that employees are expected to experience, engage in, and tolerate” (Dellinger & Williams, Reference Dellinger and Williams2002, as cited in Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011, p. 653). When reviewing these findings within a sexual-harassment-as-sexism framework, it is reasonable to assume that the antecedents to gender harassment are equivalent to those of sexual harassment. If employee perceptions of organizational tolerance of gender harassment are high, experiences of sexual harassment as gender harassment may be more frequent. Similarly, job-gender context may be a predictor of sexual-harassment-as-sexism, as gender harassment occurs most frequently in organizations where there is an unequal gender ratio.

This has some important implications for intervention work as well. Training transfer from an intervention may not be successful in organizational environments that foster gender harassment. Regardless of the content of the training, participants may be limited in their outcomes of learning in an organization with high tolerance of sexual harassment (Perry, Kulik, & Field, Reference Perry, Kulik and Field2009). For example, individuals’ perceived organizational tolerance of sexual harassment may affect trainee pretraining motivation as well as the extent to which the training transfers to the workplace (Roehling & Huang, Reference Roehling and Huang2018). As such, training transfer does not appear to be successful in organizational environments that foster gender harassment.

Training transfer from a gender harassment intervention will be most successful in an environment that suppresses gender harassment. Organizational environments that best suppress gender harassment are supportive and cooperative, with minimal gender salience and inequality, and a zero-tolerance policy for harassment (Berdahl & Raver, Reference Berdahl, Raver and Zedeck2011). For example, greater gender equality increases the likelihood that employees will confront sexist behaviors when they occur (Buchanan, Settles, Hall, & O’Connor, Reference Buchanan, Settles, Hall and O’Connor2014). Likewise, Roehling and Huang (Reference Roehling and Huang2018) note that training transfer seems to be most successful in organizations that promote empathy toward targets of sexual harassment. In addition to low tolerance and supportive environments, it also seems that the communicated reason for sexual harassment training impacts the effectiveness of the training such that organizations that conducted sexual harassment training for strategic reasons rather than legal reasons yielded higher individual perceptions of training effectiveness (Perry, Kulik, Bustamante, & Golom, Reference Perry, Kulik, Bustamante and Golom2010). Organizational environments that best suppress gender harassment will allow greater opportunity for training transfer. As such, we suggest that organizations be aware of their climate when conducting an intervention.