Understanding what harassment training does and does not accomplish is the starting point for improving sexual harassment training; “The State of Sexual Harassment Training in the #MeToo Era” shared insights from a content analysis of trainings spanning 36 years. Here, Learning Solutions presents ideas that could make sexual harassment training more effective, in part by attempting to shift attitudes, change behavior, or foster healthier corporate cultures.

Elizabeth Tippett, an associate professor at the University of Oregon School of Law, included several suggestions in her content analysis of sexual harassment training; Guild Master Julie Dirksen, a behavior change expert and principal of Usable Learning, approaches the topic through the lens of an instructional designer.

A key finding of Tippett’s study was that the framing of sexual harassment has shifted. Where older trainings presented harassment as both an abuse of power and a drag on a business’s bottom line, “the power-based frame is largely absent in more recent trainings, subsumed by the business justification,” she wrote.

At the same time, the examples used have not changed significantly, portraying demeaning and harassing sexual behavior that can seem caricatured. One example features a boss cajoling, then commanding, a support staffer to join him and a client for drinks and to “wear something nice.” But, as the recent wave of #MeToo stories has demonstrated, even such obvious quid pro quo harassment is still devastatingly common.

According to Dirksen, in cases where the harassment—or assault—is unambiguous, the issue might be a culture of “Oh, it’s just so-and-so being so-and-so,” Dirksen said—a dismissive response that managers in some recent high-profile cases have acknowledged. When high performers get a pass on harassing behavior, that is less a training issue than a problem with workplace culture.

A tougher problem with harassment, Dirksen said, is that “there’s this huge area of body language and innuendo and power dynamics and things that are not explicit and are not easy to talk about.” When addressing these “edge” behaviors in an eLearning format, which “deals best with absolutes,” and multiple-choice questions, Dirksen said, we run into problems creating effective training.

To further complicate things, context is essential to determining whether some behaviors rise to the level of harassment. Different workplaces have different rules, she said—what’s acceptable in entertainment or comedy might not be tolerated in an insurance office, for example. And the specific context of an incident matters—tone of voice, power dynamics, and more.

Approach harassment training differently

Tippett and Dirksen both said that different approaches to training could address some of the problems endemic to existing sexual harassment trainings by changing workplace culture and dynamics. Their suggestions for improving sexual harassment training include:

  • Conduct implicit-bias training: Dirksen described the benefits of learning to recognize your own biases. “Because of the nature of bias—it’s implicit and invisible to you—it’s really hard to see where you are biased.”
    How it might help: Much objectionable behavior is what Dirksen called “edge” behaviors—not explicit quid pro quo harassment but language or humor or other behaviors that could create an uncomfortable or hostile work environment. Training that helps employees become aware of their biases and avoid problematic language or behavior could improve the work environment.
  • Bring diversity training and harassment training back together: Tippett’s study found that most harassment training does not address the “closely intertwined” issue of discrimination.
    How it might help: “If you’re covering harassment, you have to talk about inclusion and equality and equity and equal opportunities for advancement because if you don’t, people walk away with totally the wrong message, which is, ‘If I just avoid interacting with everybody else, I won’t get into trouble.’ But you might get into more trouble from a discrimination standpoint,” Tippett said. Diversity flourishes—and performance improves—in an environment where people experience psychological safety. Google researchers found that the most successful teams are those where everyone has a roughly equal chance to speak up and be heard, and where team members have high “social sensitivity” to tacit cues. Both diversity and implicit bias training could improve these dynamics.
  • Use authentic content: One of the most effective examples Tippett reviewed featured restaurant workers relating their experience of harassment. Managers then said what they’d do if a similar situation arose in their restaurants, and lawyers provided feedback on the managers’ suggestions. Tippett suggests using testimonials from people who’ve experienced harassment, as well as admitted harassers, and personalizing content to learners’ “attitudes, beliefs, or self-reported behavior.”
    How it might help: Part of the problem with existing training, Dirksen said, is that “we take the personalities out of it. If we’re role-playing or we’re reading scenarios, we don’t have any emotional attachment to these people.” Real stories, where a boss—or revered celebrity—manipulates a power differential and puts an employee or applicant in an extremely uncomfortable position, are more relatable. Paired with advice on how to respond, these authentic examples can prepare people to confront likely situations.
  • Measure what’s happening: Tippett calls on instructional designers to leverage the data tracking and measurement capabilities present in most eLearning to gather data on what’s working—and to identify problem areas.
    How it might help: Much behavior at the root of discrimination is unintentional, possibly related to implicit biases. Dirksen cited research on how men and women are socialized to approach negotiation differently—and are also perceived differently when offering the same negotiating behaviors. These differences can contribute to women receiving lower pay than men in the same jobs, she said, but the behaviors are hard to measure. A neutral way to measure this would be automatically collecting data on job offers; if data show that a company is making disparate offers to men and women, the company can correct the imbalance.
  • Provide bystander training: Tippett cited Mary Gentile’s book, Giving Voice to Values. Dirksen talked about teaching people to object to harassing or troubling behavior without castigating the person responsible, referencing “guilt vs. shame.” Both describe methods of teaching employees to speak up about problematic practices or cultures without being confrontational—and by recognizing that, even when some employees’ behavior is offensive, they still share common values.
    How it might help: From her work on behavior change, Dirksen recognizes that speaking up to a boss or other higher-up can be “the most painful and awful thing,” whether the problem is doctors not washing their hands properly, a foreman asking construction workers to do something in an unsafe manner—or a boss making sexual comments or requests. Gentile’s book coaches people on “how to use all the interpersonal skills that you already have to talk about shared values,” Tippett said. Teaching learners to appeal to common values in ways that are comfortable for them can empower employees while reducing workplace conflict.

Improving sexual harassment training also requires making it more engaging than eye-roll-inducing. Tippett issues a challenge: “Do we have to talk about this as being a productivity drag? Are there other ways to frame what we’re doing that will be more compelling than ‘harassment is going to cost your business money’?”


Duhigg, Charles. “What Google Learned From Its Quest to Build the Perfect Team.” New York Times. 25 February 2016.

Gentile, Mary. Giving Voice to Values: How to Speak Your Mind When You Know What’s Right. New Haven, CT: Yale University Press, 2010.

Tippett, Elizabeth. “Harassment Trainings: A Content Analysis.” Berkeley Journal of Employment & Labor Law (forthcoming). 2018.