Passers-by on the UCLA campus write messages of support for sexual assault victims. (Evelyn Hockstein for The Washington Post)

How widespread is the problem of sexual assault in college?

The headline finding from a new national poll — that 20 percent of young women say they were sexually assaulted in college — spurred fresh debate about the question, which is a central discussion point on the nation’s campuses.

The telephone poll of current and recent students at four-year colleges was conducted by The Washington Post with the Kaiser Family Foundation. It surveyed 1,053 people ages 17 to 26 found through a random national sample. The margin of error was plus or minus 3.5 percentage points.

Skeptics say the poll was designed to overstate the problem of sexual assault in college. Other analysts say the 1 in 5 estimate probably undercounts the number of victims.

[1 in 5 college women say they were violated]

Last year a White House task force report on the issue asserted in the first line of its executive summary that “one in five women is sexually assaulted in college.” That estimate was based on earlier research, including a federally funded survey of students at two unidentified public universities.

“I felt pretty confident we were in the ballpark,” Valerie Jarrett, a senior adviser to President Obama, said in an interview last week with The Post. But Jarrett, who chairs the White House Council on Women and Girls, sought to play down the numbers question, and she declined to comment on the specific Post-Kaiser findings. She said even one sexual assault victim is too many. “No one’s denying it’s occurring,” she said. “Let’s focus on what the solutions are and not debate the statistics.”

But some do want to debate them.

An article in The Weekly Standard, a conservative journal, dismissed the findings of the Post-Kaiser poll. Its headline: “More college rape hype — this time from the Washington Post.” (The Post-Kaiser poll’s findings were about the prevalence of sexual assault — which includes a range of unwanted sexual contact — not rape alone.)

The authors of The Weekly Standard’s article, K.C. Johnson and Stuart Taylor Jr., argued that the poll “misleads readers — most of whom surely will assume that ‘sexual assault’ means criminal sexual assault — by using that criminally charged phrase for shock value in the articles while deliberately avoiding it in the survey questions.” The questions, they wrote, “are so broad as to invite survey respondents to complain about virtually any encounter that they later regretted, including many that were not sexual assault or rape as defined by law.”

Johnson and Taylor said a more reliable source is the federal Bureau of Justice Statistics, which in 2014 reported that female students suffer rape and sexual assault at an annual rate of 6.1 incidents per 1,000 women.

A National Research Council report last year concluded it is likely the BJS survey method undercounts rape and sexual assault.

The Post-Kaiser poll’s script covered five types of “unwanted sexual contact or sexual assault”: forced touching of a sexual nature, oral sex, sexual intercourse, anal sex and sexual penetration with a finger or object. It also provided more information about those scenarios if a respondent wanted clarification.

After they were read that preface, poll respondents were asked if they experienced unwanted sexual contact involving force or threats of force, or if they had such experiences while they were unable to provide consent or stop what was happening because they were incapacitated.

Those who said yes to either question were classified as sexual assault survivors.

[Numbers and methodology: the Post-Kaiser poll results]

Mary P. Koss, a professor of public health at the University of Arizona, who has studied sexual assault for decades, said the Post-Kaiser results were broadly consistent with what other research has found. But she said she wished the poll’s wording had included more detail in asking about specific behavior that constitutes sexual assault. “I don’t think you overestimated,” she said. “I think you underestimated. In a way, if you’re going to make a mistake, that’s the one to make.”

Post reporters also interviewed dozens of poll respondents about their experiences with sexual assault, and they described a wide variety of incidents, from groping to unwanted intercourse while intoxicated to forcible rape.

[Dozens of survivors tell their stories]

Other efforts are underway to define the problem. The Association of American Universities organized a survey given this year to hundreds of thousands of students at more than two dozen prominent universities. Many other colleges are surveying students on the issue too.

Also, the National Institute of Justice, a unit of the Justice Department, is preparing a journal article examining studies of the prevalence of campus sexual assault.

“We wanted to present the breadth of statistics on this topic to show there are varying ways of measuring sexual violence, but the results are generally similar,” the institute said in a statement. “Regardless, we also wanted to discuss that although statistics are helpful, sexual assault is occurring and responses to this issue should not be guided by numbers.”

That is Jarrett’s point. The Obama administration has sought to raise the profile of the issue for several years, through civil rights investigations of colleges and aggressive use of the bully pulpit. Last fall, President Obama and Vice President Biden launched a public relations campaign for prevention of sexual violence called “It’s on Us.” Hundreds of campuses are participating.

“This problem is ubiquitous,” Jarrett said. “It’s not just a handful of college campuses that have this problem. It’s everywhere.”

Jarrett added: “We’re doing everything we can within the federal government.  … But ultimately cultures change when the people who make up the community decide it’s time to change.”

Read more:

1 in 5 college women say they were violated

‘I woke up. He was in the room. I didn’t know who he was.’

Dozens of survivors tell their stories

The meaning of consent