Over the past year, the Task Force on the Prevention of Sexual Assault has been gathering information to lay the groundwork for interventions to lower the incidence of sexual assault, harassment, and other misconduct at Harvard, as well as for improvements in supporting students who have experienced such misconduct.
This week, the task force is launching its most far-reaching effort yet, sending out a Web-based survey to 20,000 members of the student body, targeting all degree candidates as part of a national effort involving 28 universities to understand the extent and nature of the problem, both on their home campuses and across the country.
The task force’s work is just the latest effort to get a handle on the problem at Harvard. Last year, the University adopted policies and procedures to address sexual assault and harassment and opened a new Office for Sexual and Gender-Based Dispute Resolution. In addition, the issue has been discussed by different Schools. Last week, for example, the Radcliffe Institute for Advanced Study sponsored a panel discussion on how government policies affect the incidence and tolerance of gender-based violence.
The Gazette sat down with the task force’s chairman, Steven Hyman, a professor of stem cell and regenerative biology and Harvard University Distinguished Service Professor, and David Laibson, the Robert I. Goldman Professor of Economics, who participated in the survey design. The two discussed the survey, what students can expect to be asked, and the potential benefit of taking a hard look at a difficult issue.
GAZETTE: The Task Force on the Prevention of Sexual Assault is sending out a survey to students. What is the task force hoping to learn and why is the survey needed?
HYMAN: First, the mission of this task force is twofold: one is prevention of sexual assault and other forms of sexual misconduct, and the other is making sure we do our best to support people who have experienced sexual assault and misconduct.
In order to prevent these unwanted behaviors, we have to know what the risk factors are that contribute to their occurrence. We also have to know how many episodes there are each semester in order to have a chance of knowing whether our interventions are succeeding.
To design effective preventive interventions for the Harvard context, the task force has undertaken three complementary efforts. One endeavor was a very intensive effort to listen within small and large meetings involving many different groups across Harvard’s schools. Stephanie Khurana took the lead for the College and provided important intellectual leadership.
A second effort was to review the academic literature to determine what was really known about assault and harassment. Lisa Berkman of the Harvard T.H. Chan School of Public Health took the lead. This effort was complemented by hearing reports from other universities, including Brown and MIT, and interviewing experts at other Universities.
The third effort was to design and implement the survey that is about to be released at Harvard. Not only is this a way to learn about our own institution, but also to benchmark ourselves against similar institutions. Among many advantages of surveying multiple institutions with the same instrument is the possibility of learning from differences and understanding best practices for intervention.
GAZETTE: How many institutions are involved with the survey and what kinds of things will you be asking?
LAIBSON: The survey is a collaboration, primarily with the AAU, the Association of American Universities, which is a consortium of 62 research universities in the U.S. Twenty-eight will be participating in the survey.
We participated in every stage of the process of creating the survey. We had a set of local conversations and convened an Ivy Plus consortium group, which is a conversation among the Ivies and a few additional schools.
There were three bids from external survey research firms, and Westat emerged with the strongest bid. Then a survey design committee organized by the AAU took over. That committee, which we participated on, met once to twice a week working with Westat to design the survey. The design process started in October and continued until just a few weeks ago.
We were the first school to pretest the survey, and we did that in January. The word “pretest” is a little bit odd because on the one hand it sounds like people actually took the survey, but we didn’t want that. We really wanted students to react to the questions and tell us whether they were clear, comprehensible, personally relevant; appropriate in every conceivable way.
GAZETTE: Can you describe the content of the survey? How long is it, how many questions, and what kind of topics do the questions cover?
LAIBSON: Length is better summarized with time than by a number of questions, because it’s conditional. If people are reporting incidents of an assault, we’re going to follow up on that. So if you say yes, that opens up more questions.
We’ve thought a lot about good survey methods and the burden on our respondents. Since we would like a high response rate, we wanted to create a survey that wouldn’t drive people away because 45 minutes into it they’ve got to go to dinner. We worked very hard to keep the survey down to 20 to 30 minutes for a typical respondent.
We also want to emphasize that confidentiality — really anonymity — is important. We’re sending people a link. When they click on that link, any identifiers — to email or to name — is severed.
HYMAN: At no time in the process does Harvard University possess any identifying data, like IP addresses. The survey is performed on the Westat site and we get data back stripped of all identifiers. Even the emails to students are sent from Westat and not from Harvard.
GAZETTE: What should students expect? And if it’s not from Harvard, how does it get through spam filters?
LAIBSON: We have been “white-listed” in the Harvard email system, but of course it cannot be white-listed in every email system. So that means if someone is forwarding all their emails to Hotmail, it could get spammed in Hotmail. So if students didn’t see it in their inbox at the end of the day on the 12th, they should check their spam filters, particularly if they’re forwarding their mail to another service.
It’ll come from Westat, they won’t see Harvard.edu. It’ll contain an invitation from Drew Faust to participate, but it won’t be from her address. There will be a link, they’ll click on the link, and at that point their identity is severed from the survey. All we will know is what we collect on the survey.
We’re going to ask them about demographics, about their living situation, many things about the climate at Harvard. We’re trying to cover a lot of territory. First we’re going to be asking about sexual harassment and, of course, about sexual assault. We’re going to be asking about intimate partner violence, and about stalking.
Sexual assault is very complex and problematic, so we’re not going to ask generic questions like, “Have you been the victim of a sexual assault?” We’re instead going to be very explicit and talk about very particular situations with very precise wording and ask, “Did this happen to you?”
HYMAN: We want to know about specific experiences and behaviors, not how people have labeled them, because labels can be understood very differently among the members of our community.
LAIBSON: Sexual assault can take many forms: through force, through attempted force, through incapacitation, so all of these different channels will be measured.
Then we’ll follow up and ask detailed questions about the characteristics of the event, the reporting, follow-up, etc. We want to get a relatively full picture of the circumstances and context of the incident, and then how the incident was or was not followed up and whether those follow-ups were satisfactory.
So we’ll have a pretty complete picture of the characteristics of each incident plus an overall measure of prevalence by category.
GAZETTE: Are there any questions aimed at perpetrators?
LAIBSON: We decided not to ask those particular questions because we were concerned about length of the survey and we didn’t think we’d get honest answers. There may be ways to ask those kinds of questions, but not on this type of a survey.
HYMAN: We think it’s very important. It’s been alleged based on some research that a small number of repeat offenders are responsible for a substantial fraction of assaults. It would be very important to know if that is true, but that would require a very different survey.
GAZETTE: Is there a concern that the survey might be painful for someone who has had a bad experience and been assaulted?
HYMAN: I would describe the survey’s language as direct and precise. If we were to focus overmuch on the risks of upsetting people, we would not be able to obtain the information we need. We would then be sacrificing the ability to make the situation better for others. So we hope that students will be poised enough to understand that even if it’s unpleasant, it’s important.
LAIBSON: I want to reiterate that point. In getting student feedback during the pretest, one of the concerns was how they reacted to the language. And across the wide swath of the student community, including activist groups, there was uniform support for the approach that we took.
There is also a link to reporting and support services directly from the survey. Anyone who’s feeling traumatized in taking the survey is encouraged to stop if they want to. Our goal is not to make people take it who don’t want to take it.
GAZETTE: But at the same time, it’s important to get an accurate view of what’s going on?
HYMAN: If we don’t get a large and representative sample, it will be very hard to know how to design preventive interventions and very hard to know whether they have been successful. It’s just of critical importance that our students participate and take this seriously.
After we design and the University implements specific, preventive interventions, we cannot assume that they will improve the situation significantly. We must be able to measure whether we have made a difference in the incidence of assault and harassment. To do that we need good baseline information and then follow-up data over time.
GAZETTE: So, in addition to talking in fairly blunt language about what happened, there are questions also about where, when, who?
LAIBSON: Yes, we address that, including alcohol.
There are also Harvard-specific categories of responses, reflecting, for example, the organization that provides support and reporting services. At Harvard it would be OSAPR [the Office of Sexual Assault Prevention and Response], which obviously wouldn’t exist at another school. We ask about OSAPR here.
There are idiosyncratic organizations or living arrangements here and maybe not at other schools, like Houses, so these response categories will vary across all 28 schools in this initial wave of the survey.
GAZETTE: How many surveys are going to go out to the Harvard community?
LAIBSON: The decision AAU and Westat made is to survey degree candidates. There are 20,000 degree candidates at Harvard. And all of them are going to be surveyed and all will be offered a $5 Amazon card if they participate in the survey.
GAZETTE: What do we know already about sexual assault here at Harvard?
HYMAN: We simply do not know the true prevalence. Our goal is to gain a good estimate of the behaviors that constitute assault and harassment and to learn how best to increase honest reporting when such events occur.
GAZETTE: What are the factors that make the issue such a black box right now? Why don’t we know?
HYMAN: Across cultures and populations, sexual assault is underreported. Very often, those who have experienced assault or harassment don’t feel safe reporting for fear of stigmatization, retaliation, or of being retraumatized by an investigation. We have students coming from cultures where women who report having been assaulted may be harmed or marginalized for life.
We understand the disincentives to reporting are deeply ingrained in many cultures, including ours. Having a technology that allows people to take a survey and have their identities truly and credibly protected is an important recent development.
LAIBSON: This is the first time that an effort like this, with national scale, has ever been conducted. The survey will also give us the ability to compare across schools and identify successes. One thing that might emerge is that we find some schools or subpopulations within schools that have much lower rates of sexual assault. And, if that’s the case, I’d like to learn more about those environments and the mechanisms, the correlates, of that success.
If we discover, just as an example, that 24 of the 28 schools have a very high rate of sexual assault but four of the 28 have a very low rate of sexual assault, the first thing I’m going to want to do is call up that community and find out what’s different. It may take a few years to get to the bottom of that, but I think this could be a very important channel for learning about the mechanisms that reduce the problem.
HYMAN: We also hope to be resurveying at intervals to learn what has happened and to know whether we are on the right track.
GAZETTE: When will results be announced?
LAIBSON: Harvard will be releasing its results during the fall semester and I believe that the national results will be released around that time.
It’s important for us to be transparent about the results. We’ll be discussing with the community both the national Westat results and the Harvard-specific results. We’ll be transparent about what’s happening here and in comparing our outcomes with the outcomes in the national sample, the 28 schools.