A Bogus Statistic That Won't Go Away   

At the end of an interview with actress Ellen DeGeneres, aired the night her lesbian television character "came out," anchor Diane Sawyer addressed viewers of "PrimeTime Live":

"And as we close, we're going to repeat a government statistic that a gay teenager is some three times as likely to attempt suicide as another teenager. Ellen DeGeneres has said whatever happens to her, tonight's broadcast was in part to hold on to them."

It's a "statistic" that's been repeated innumerable times. The trouble is, there is no scientifically valid evidence that it's true. In fact it is not a government statistic at all, but rather the interpretation of a social worker.

So how did it come to be broadcast to a national television audience? Sawyer's office will say only that Sawyer got the number from DeGeneres.

But Sawyer is not the only journalist who has reported the suicide figure uncritically. Had she done a quick Nexis search, she would have found any number of newspapers — from the New York Times to the Chicago Times to the Los Angeles Times — quoting the "statistic" that gay and lesbian teens are two to three times more likely to attempt suicide and that they may account for 30 percent of the 5,000 youth suicides committed in the United States every year.

This last number is the real attention-getter — often played up in drop quotes, graphics and cutlines. But according to Peter Muehrer of the National Institute of Mental Health, "There is no scientific evidence to support this figure." Nor is there any to support the number of suicide attempts.

The real story is not at all as clear or dramatic as these figures imply.

A panel convened in 1994 — with representatives of the U.S. Department of Health and Human Services, the Centers for Disease Control and Prevention, the National Institute of Mental Health, the American Psychological Association, the American Association of Suicidology, and gay and lesbian advocacy and service groups — made this finding: "There is no population-based evidence that sexual orientation and suicidality are linked in some direct or indirect manner."

Joyce Hunter, the immediate past president of the National Lesbian and Gay Health Association, participated in that 1994 meeting. Three years later, she says the data are still unclear.

Yet advocates for gays and lesbians still provide journalists with these numbers — and journalists still report them. And not without consequences.

The stunningly high number of suicide attempts and suicides represented by these figures has shaped public perceptions of gay teens for nearly a decade. Typically they are portrayed as emotionally vulnerable and as society's victims, even though Hunter agrees with mental health researchers that most gay and lesbian teens, like teens overall, are emotionally resilient people who "go on to develop a positive sense of self and who go on with their lives."

But the conventional wisdom that gay teens are killing themselves has become so ingrained that it can influence public policy. Gov. William F. Weld of Massachusetts, for instance, cited the suicide "statistic" as a driving force behind his state's creation of a Governor's Commission on Gay and Lesbian Youth.

Muehrer and his colleagues worry that the public focus on gay teen suicides might contribute to a phenomenon called "suicide contagion," in which troubled gay teens might begin to see suicide as an acceptable way out of their identity struggles.

According to Howard Bray, director of the Knight Center for Specialized Journalism at the University of Maryland College of Journalism, the gay teen suicide figure illustrates an important lesson for journalists. He says it is critical to remember that, when it comes to statistics, journalists face a "minefield."

Victor Cohn, author of "News & Numbers: A Guide to Reporting Statistical Claims and Controversies in Health and Other Fields," suggests a remedy. "In a nutshell," Cohn says, "the most important question to ask anybody is, 'How do you know?' "

If reporters had asked that question about the gay teen suicide rate, they would have learned that the "two to three times more likely to attempt suicide" and the "30 percent completed suicide" figures emerged from a 1989 U.S. Department of Health and Human Services task force report on youth suicide. The numbers were drawn from a single essay written, at the department's invitation, by a San Francisco social worker named Paul Gibson.

Gibson did not base his numbers on original research, but rather on his own interpretation of available literature — literature that has since been criticized for methodological weaknesses. He concluded that gay and lesbian teens were two to three times more likely to attempt suicide, and then went on to extrapolate that such a figure also would indicate that gay and lesbian teens may account for 30 percent of all teen suicides. Mental health researchers, however, consider suicide attempts and actual suicides to be distinct phenomena.

Although packaged in a government report, Gibson's paper was not government research, nor were his figures government statistics. But they were quickly picked up and presented as such by gay advocates and many reporters.

Gibson says he never intended for the 30 percent figure to be used as a statistic, but he stands by both numbers. "I think that there is ample evidence available to indicate that these youth are at risk and that we need to help them now," he says. "We can't wait to help these youth until complete information is available."

But reporters can certainly inform their readers when accurate information is elusive.

In such a case, Cohn suggests, reporters should be honest with their audience. If the study is tentative, say so. His general rule is this: "Understand that all studies are not equal, that just because someone said they've done a study or looked at something or made a count, that doesn't mean it's correct or worth possibly believing."