Asking Too Much, or Not Enough?

Some colleges conducting climate surveys are criticized by sexual assault victims and advocates for the explicit questions. But researchers say such language is key.

May 27, 2015

This spring, institutions that administered the campus sexual assault survey created by the Association of American Universities found some unexpected critics: students who had been assaulted and their advocates on campus.

Advocates have long told colleges that they should conduct anonymous “climate surveys” to better assess the prevalence of sexual assault on their campuses. When the University of Michigan began soliciting responses to both the AAU survey and one of its own creation, some students said the language used in the survey's questions made them uncomfortable or even triggered dark memories of their assaults.

The questions asked students if they had ever experienced a number of specific sexual activities without their consent, describing those actions with words and phrases such as “oral sex” and “penetration,” and defining the terms using definitions such as “when a person puts a penis, finger or object inside someone else’s vagina or anus.”

The complaints raise a quandary for colleges hoping to better estimate how prevalent sexual assault is on a given campus. Consensus among researchers is that using more specific language in climate surveys results in more accurate data about instances of sexual assault.

But if that specific language is driving victims of sexual assault away from the survey, could that, too, cause cases of sexual violence to be undercounted?

"The more behaviorally specific, the more accurate the findings," William Flack, an associate professor of psychology at Bucknell University, said. "Thus, asking about 'sexual assault' or 'rape' is not recommended because respondents often do not know how these terms are defined by their institution or the law. Furthermore, it may be difficult for respondents to acknowledge that they've been assaulted or raped because most perpetrators are acquaintances or friends."

The AAU announced in November that it had contracted with a research firm, Westat, to develop and carry out the campus climate survey for any of its member institutions that wanted to participate. The association said that one goal of the project was to fend off congressional efforts to require universities to annually survey their students about the prevalence of sexual assault, a move opposed by many in higher education.

At the time, the AAU survey was criticized by several dozen scholars who study sexual violence as well as some victims’ advocates for, among other things, not pledging to release campus-by-campus data and to share only aggregate data. Critics also said the process lacked transparency and input from enough scientists who study sexual assaults on campuses.

Only 27 of the 60 AAU member campuses decided to participate in the survey. Some of those institutions do plan on releasing individual survey results, they said, but not until this fall. The aggregate data will be released by the AAU around the same time.

Not all of the colleges that participated in the survey heard complaints from students, college officials said, but criticism that the language in the surveys could be potentially triggering did arise at a handful of institutions in addition to Michigan, including Harvard University and the University of Pennsylvania.

"Some of the language used in this survey is explicit and some may find it uncomfortable, but it is important that the questions are posed in this way so that students are clear about the meaning," the University of Michigan said in a statement. "In order to understand the climate here, we need to ask direct questions about topics that some may find sensitive. It is only by directly collecting this information from students will we be able to prevent negative experiences and effectively respond when they do happen."

At Penn, some students also complained that they didn't realize the survey was about sexual assault, as it was referred to as a "climate survey," as these kinds of surveys commonly are. Thinking the survey was about climate change, the students claimed, they deleted the email. The university declined to comment on the complaints, and said it does not plan on releasing its response rate until the fall. Harvard had a response rate of 52 percent, thanks in part to a large ad campaign on campus, including a video message from Harvard graduate Conan O'Brien.

The phrasing in the surveys has also been criticized -- especially in the conservative press -- for conflating what critics considered to be more minor transgressions with sexual assault.

"Sexual assault and sexual misconduct refer to a range of behaviors that are nonconsensual or unwanted," the survey states. "These behaviors could include remarks about physical appearance or persistent sexual advances. These could also include threats of force to get someone to engage in sexual behavior such as nonconsensual or unwanted touching, sexual penetration, oral sex, anal sex or attempts to engage in these behaviors."

In an opinion piece for The New York Post, Naomi Schaefer Riley, a senior fellow at the Independent Women’s Forum, mocked Harvard's survey’s trigger warning and said that despite its specific language, the survey doesn't really differentiate between various kinds of sexual misconduct.

“So which is which?” she wrote. “What is sexual assault? What is sexual misconduct? What is harassment? And is there any category for simply obnoxious behavior? Indeed, there are questions about people tweeting offensive sexual remarks thrown in with questions about anal penetration. And, oddly, nowhere in the whole document does the word ‘rape’ appear. Perhaps that’s because rape is a word that respondents might be a little more careful about using. But the school isn’t trying to understand a problem. It’s trying to cover its behind.”

While the introduction covers many behaviors, the survey includes specific questions about whether students have experienced specific forms of gender-based misconduct -- including in-person and online harassment -- and sexual assault, such as nonconsensual intercourse, sex while unable to consent due to intoxication and so forth.

Barry Toiv, vice president for public affairs at AAU, said how the results will be presented in the association's final report is still being determined, but that “the survey was done in this way in order to gain as accurate and specific an understanding as possible of what is occurring at these universities.”

Indeed, while the wording in the AAU surveys might make some students uncomfortable or stir up unwelcome memories among victims, the growing consensus among researchers is that using “behaviorally specific language” is the best method for more accurately understanding the prevalence of sexual assault.

That consensus has led the Bureau of Justice Statistics to second-guess its own findings. Last year, it asked the National Research Council to look into the whether its studies based on the National Crime Victimization Survey were undercounting cases of sexual assault. The council’s conclusion: by using “ambiguous” words and phrases like “rape,” the bureau is likely undercounting rape and sexual assault. Studies have repeatedly shown that many young women and men who are survivors of rape and sexual assault have trouble identifying it as such.

“Self-report questions with behaviorally specific language have been used for more than 30 years in over 1,000 studies with no serious problems,” said Mary Koss, a professor of public health at the University of Arizona. “There are actually research studies documenting no increases in distress [for victims] and many welcoming the opportunity to report these experiences.”

Koss said that it’s important to keep in mind that victims of sexual assault might be experiencing post-traumatic stress disorder, and that surveys should always be clearly labeled and optional. But fear of triggering symptoms of “PTSD re-experiencing” among survivors, she said, shouldn’t stop researchers from using current best practices for determining the prevalence of sexual assault on college campuses.

“Scientists after many years of study are in consensus that these types of questions are the best way to measure sexual assault,” Koss said. “‘Triggering’ is disempowering language that has been interjected into the dialogue without thought of its unintended consequences on women’s confidence in their ability to survive and thrive.”


Back to Top