The Delicate Business of Conducting Sex Surveys : Honesty: Contradictions between the results of a new poll and earlier studies prove at least one thing--we don’t always tell the truth about our most intimate moments.
Ever since the 1953 Kinsey Report, conventional wisdom has held that 10% of the U.S. population is gay. Forty years later, the Janus Report stated that 22% of American men had had a homosexual experience, but last week the Alan Guttmacher Institute reported that figure is only 2.3%.
Likewise, the public has heard conflicting reports on infidelity, the practice of safe sex and the prevalence of rape.
Does anyone really know the truth?
Knowledge about sexual behavior has become an urgent matter in the United States, with widespread concern over AIDS, abortion and teen-age pregnancy. But while sex surveys have become more common, sophisticated and accurate than ever before, the art of turning intimate acts into scientific data remains . . . delicate.
Today researchers can ask difficult personal questions that would have been unthinkable in years past, such as “Have you had anal intercourse? With men? With women?” But they acknowledge such questions can also skew a survey by either repelling or attracting participants, or embarrassing people into fudging.
“You always wonder: Are you getting over-reporting, or under-reporting?” said Kristin A. Moore, executive director of Child Trends, a Washington, D.C.-based research organization. With the exception of questions regarding abortion, which can be independently verified, “you don’t have a way to tell.”
Scientifically valid studies of sexual behavior, with samples that represent all groups, did not begin until the mid-1980s. So far, there have been only a half-dozen “good studies,” said Tom Smith, director of the General Social Survey at the University of Chicago’s National Opinion Research Center. The General Social Survey has been tracking trends in American society since 1972.
Pioneer sex researcher Alfred Kinsey published “Sexual Behavior in the Human Male in 1948” and “Sexual Behavior in the Human Female” in 1953 based on face-to-face interviews with 5,300 men and 5,940 women, all Anglos, in the United States and Canada chosen by “quota sampling and opportunistic collection.”
The “Janus Report on Sexual Behavior,” conducted between 1988 and 1992 by sex counselor Samuel Janus and his wife Cynthia, a physician, used written questionnaires and 125 interviews with 2,765 men and women, chosen to approximate a cross section of the United States.
In 1991, the Alan Guttmacher Institute’s National Survey of Men studied 3,321 men, ages 20-39, in a population-based, nationally representative survey.
Today’s researchers believe their numbers are more reliable than Kinsey’s, or those that appear in men’s and women’s magazines, because they try to systematically account for geographical, ethnic, social and economic variables. Most of the current research on sexual behavior involves face-to-face interviews--and interviewers are trained extensively. And even though the sex researchers have only a fraction of the experience of, say, unemployment researchers in refining their questions, they’re working on improvements.
One of the major hurdles they encounter is honesty.
When people understand the medical or public health reasons for a sex survey, most of them respond honestly, researchers said. But some do not.
By checking reports from hospitals and clinics, researchers know women under-report their abortions by as much as 50%, Smith said. Most assume the women are afraid the interviewer will think less of them if they admit to having had one or more abortions. Other researchers speculate the women are afraid word will leak out to their friends or family. Some wonder if the women don’t even want to remember it themselves.
Homosexuality is also thought to be under-reported, but no one knows by how much.
Researchers suspect single men may over-report how often they use condoms because they want the interviewer to think they are socially responsible, Smith said.
Other researchers believe some men tend to exaggerate or “round up” the number of their sexual experiences.
If you ask men and women about sexual partners, men report more than women, said Kathryn London, a demographer with the National Center for Health Statistics in Hyattsville, Md. “Theoretically, the number of partners should average out.”
But Smith said his research shows the opposite. He asked groups of husbands and groups of wives how often they had sex, and the wives said they had sex more often than the husbands.
Particularly troublesome to social scientists are questions of infidelity, rape and sexual harassment, Smith said.
It has been widely reported that one in three women have been raped. Smith said official crime statistics of 2 to 4% are “clearly way low,” but one-in-three has not been scientifically established.
Not only might different people have different definitions of sexual encounters, but their answers vary widely depending on how the question is asked or who is asking, researchers said. One reason Kinsey’s 10% on gays figure is off, Smith said, is that the question was limiting. Kinsey researchers asked whether men were exclusively homosexual for three or four years as an adult. “He has another figure, which is only 4%, when asking if they were exclusively homosexual” over their lifetime, Smith said. “He has other numbers, which are higher, in the 30s, asking whether they ever had any homosexual experience at all.”
The Janus Report asked, “Did you ever have a homosexual experience?” Smith said. “Who knows what people have in their minds? If you’re a straight guy and a homosexual made a pass at you in a bar, do you count that?”
Since adultery is technically illegal in many states and opinion polls show 75% of people say it’s always wrong, Smith said, “We would never ask, ‘Have you been unfaithful?’ ” Instead, he said, researchers ask for the number of sex partners in the last year, then later cross check for marital partners. If the number of partners exceeds one, they can chalk up one infidelity without having to ask.
In an effort to encourage people to talk more frankly and truthfully about their sexual behavior, researchers are experimenting with a variety of techniques.
Interviewers are often trained to put people at ease, assure participants of confidentiality and ask questions in a “value-neutral” way, said John O.G. Billy, senior research scientist at Battelle Human Affairs Research Centers in Seattle, which conducted Guttmacher’s National Survey of Men.
“You try not to convey the message that it’s wrong or right to do this in any way. Mostly you ask them very matter of factly, ‘Have you ever engaged in anal intercourse?’ ” Some interviewers are trained to avoid body language that would indicate disapproval.
Sometimes definitions of the terms oral, anal or vaginal intercourse are placed on cards in front of the respondents to ensure everyone is talking about the same thing.
The National Survey of Men used women to conduct the standard face-to-face interviews. At least one study indicated people are more willing to divulge personal sexual behavior to women than to men, but researchers are unsure of how talking to women affected the reliability of the men’s answers.
Occasionally, frauds pose as researchers for prurient reasons, but genuine researchers say they take confidentiality seriously. For instance, Moore said she once had a professor who was willing to risk jail rather than divulge the identity of a sex survey participant. In that case, a student interviewer had been killed and police wondered if the murderer might be someone she met while conducting the survey. The professor refused to divulge her records, but was never called to court since the killer was found elsewhere.
Researchers sometimes use laptop computers or self-completion forms so that people can answer the questions alone. They can mail the survey back to an office where it is opened by someone who knows nothing about the respondent. “They are just asked numbers: three partners, five times a week, whatever,” Smith said. “That is well demonstrated to greatly improve accuracy and truthfulness. (Still), it doesn’t guarantee everyone tells you the truth.”
Researchers said the worst surveys are those conducted by men’s and women’s magazines, although their figures may be picked up in the general press. Those surveys are inherently biased by the readership and the voluntary nature of the participants. “In case of sexual behavior samples, you get a profile much more sexually active and adventurous than the average citizen,” Smith said.
In some cases, the biases are “enormous,” Smith said. Consider the 1976 Hite Report, which concluded that 75% of married women were unfaithful to their husbands. Author Shere Hite did not use a representative sample; she mailed out 100,000 surveys and about 4,000 were returned, a rate well below that of a well-conducted study. “No other study gives a number anywhere near (the 75%),” Smith said. The highest number is 50% even in other “bad studies,” he said.
Researchers believe their sex studies are becoming increasingly accurate because they are starting to obtain similar results with similar methods.
For instance, although the National Survey of Men’s findings surprised many (and angered some), they were supported by four previous recent studies in Europe and the U.S. Smith said. His own figures show there are very few “lifetime homosexuals,” he said.
These new studies suggest that 93% of the 98% of the adult population that has had sexual experience is exclusively heterosexual, 6% is bisexual, and 1% is exclusively homosexual, he said.
