David Streitfeld

Articles by Subject > By Publication > Controversies

Shere Hite and the Trouble With Numbers

For the famed sex researcher, questions and controversy over "Women and Love."

The Washington Post
November 10, 1987

Ever since Time magazine put her on the cover a month ago, Shere Hite has had trouble getting her message across. Never mind the great publicity for her new book, "Women and Love"; Hite felt that Time unfairly quoted unsympathetic critics who hadn't read her 922-page tome.

Shere Hite is a woman with an important message: Behind her alarming figures, she says, lies "a huge ideological transformation," with women redefining their lives and demanding more from their men.

But people aren't listening.

Instead, they're focusing on the surprising statistics (84 percent of respondents to her questionnaire are not satisfied emotionally with their relationships; 95 percent report "emotional and psychological harassment" from their men; 98 percent desire more communication; just 13 percent of women married more than two years are "in love").

Instead, they're saying that only unhappy women are likely to take the time to answer her 127 essay questions, and thus her sample is representative only of the discontented.

Instead, they're talking about her own quirky behavior, like the time two weeks ago when she reportedly slugged a limo driver who called her "dear." The next day, she stormed off a TV interview that attempted to confront her with the chauffeur.

Hite is always on guard against trivialization of her work. "The book is an intricate philosophical discussion by thousands of women who are discussing their emotional lives," she says. " . . . I guess it would be expecting too much to get the press to talk about philosophical issues."

No wonder she's sometimes driven to take extraordinary measures to convey her ideas. Now a lawyer screens her interview requests. Indeed, she has had a number of unpleasant experiences with the press.

Schedule on your TV show an expert who might pose questions about Hite's methodology and Hite threatens to cancel. Compliment her getting the cover of Time before your live radio show airs and it ends up in an argument and her departure moments before broadcast. Call her a charlatan and she sues for $15 million.

And ask her to prove that an employee of hers really exists -- an employee who has made phone calls and written letters to reporters on her behalf, but whom no one except Hite and her husband appears to have seen -- and she throws you out of her apartment.

The plan was, "Women and Love" would be unveiled at an Oct. 5 press conference. Friendly academics would be on hand to salute Hite and her work, and she could present her conclusions in a controlled setting.

She didn't do that six years ago, with "The Hite Report on Male Sexuality." "The debate never got off onto a good level in that book," she says, "probably because I didn't write out a press synopsis and all that stuff."

This time she did all that stuff. And she wanted to make sure people came. A woman who identified herself as Diana Gregory from Hite's office called The Washington Post nearly three weeks before the press conference. Gregory asked the reporter who had requested an interview with Hite if he wouldn't really rather come to the press conference instead.

For 50 minutes, she quizzed and argued. The next day one of the publicists on the book phoned the reporter to apologize. Gregory had no right to make a call like that, the publicist said; and Hite herself was willing to write a letter of apology.

Several days later, another Post reporter got a letter from Diana Gregory, asking him to attend the press conference. And Gregory also contacted The New York Times.

Then, on Oct. 5, Time magazine's cover story on Hite appeared. A number of people who had heard of the disruptive Diana Gregory were surprised to learn from the story that Hite's original name was Shirley Diana Gregory.

During an interview in a Washington restaurant, Hite said the similarity in names was just a coincidence. Diana Gregory had worked for her for about four years, she said, "doing everything there is to be done . . . answer the mail, write checks, go run and get things, cook lunch, retype things, take them over to Knopf."

William Loverd, head of publicity for Knopf, says: "Diana Gregory has never been known to appear in our offices. We've never met her." Hite's editor says the same thing.

The same is true of independent publicity people who worked with Hite. "As far as I remember, I never met, never talked to on the phone anyone named Diana Gregory from Hite's office," says independent publicist Cathy Saypol. "I have not in any of the times we worked on this project had any contact with a Diana Gregory," says independent publicist Karin Lippert.

In the acknowledgments of "Women and Love" -- in which she thanks more than 40 people -- Hite didn't mention her employee of four years, Diana Gregory. "No, I didn't, did I?" she said. She said she forgot. "The last parts are done at the last minute."

The similar voices? "In the Midwest" -- Hite was born in St. Joseph, Mo. -- "a lot of people sound the same."

There's also the handwriting. The signature on Diana Gregory's letter to The Post was compared with Hite's own signature by handwriting expert Charles Hamilton. "These are the same handwriting," he said.

Asked about the identity of Diana Gregory in the Washington interview, Hite responded: "Why don't you come up and visit? That's the best way to squelch your curiosity." She offered to line up her employees -- three full-timers, she said, and one part-timer -- for presentation.

At Home With Hite

Hite lives in an elegant, roomy Fifth Avenue duplex apartment with her husband, pianist Friedrich Horicke. The office/living room section is an open area -- partially cleared this day so an interview with Brazilian television can take place.

Horicke is tall, amiable, nattily dressed. "Yesterday the German consul told me I was spoiling the prices," he says in his heavily accented English. The prices? "Maybe it's funnier in German," interjects Hite. He says he means that, as Hite's perfect man, he's spoiling "the male market."

He was only 13 in 1976, when the first Hite report appeared -- the one that sold the most copies and made her name. (By citing testimony from many women who said they could not achieve orgasm from the simple act of intercourse, that book helped redefine women's sexuality.) Do people get confused, call him Mr. Hite? "Some do," he says. Says Hite, 44: "I call him Mr. Hite every day."

"While she's bashing me," says Horicke, and laughs. "Some people in the media have called the book a male-bashing book," he explains. "That's why I'm saying I'm a bashed male." He laughs again. "Joke!"

On a couch, a woman is sitting by a phone. "This is Diana," Hite says. A couple minutes later, Diana makes another call, inviting people to Hite's Halloween party. "Hello, this is Diana Gregory," she says, apparently into an answering machine. "I'm calling on behalf of Shere Hite . . . " She pronounces it "Sherry"; actually, Hite pronounces her first name "Share."

Gregory says she can't talk to the reporter because she has more calls to make, but, in another few minutes, she tells Hite she has to leave. A reporter's request to see her driver's license or some other proof of identity doesn't make it very far.

"We're not living in a police state yet," Hite says, quite loudly. Horicke, all humor gone, announces at full volume, "I think you should get out!" The reporter is quickly whisked out the door. Through it all, the Brazilian television crew watches, dumbfounded.

The Numbers of Love

Whatever the story with Diana Gregory, public attention and criticism have focused on the statistics in "Women and Love," which is subtitled "A Cultural Revolution in Progress."

Through an exhaustive seven-year research project, Hite says, she has gathered her data for this, the third Hite Report. Seventy thousand copies are now in print. "Because I went for statistical representativeness," she says, her results "could be replicated" on a much larger scale with no more than "a 10 percent range of variation" in the results.

If she had gotten back millions of completed questionnaires instead of 4,500, Hite says, the same percentages of women would likely have answered the same way. "No sample is perfect," she says, but "I think it probably would be quite close."

This assertion of broader significance -- that vast numbers of women are, indeed, "suffering a lot of pain in their love relationships with men" -- is responsible for the book receiving such wide attention. As the "Women and Love" jacket puts it: This is a "startling revelation of how women in America today feel about love."

"In my view, that's going a little too far," says John L. Sullivan, a political science professor at the University of Minnesota. A survey researcher and methodologist, Sullivan read the manuscript of "Women and Love" and gave Hite a favorable evaluation, which is quoted in the book.

Hite "does slip into" claims that her respondents mirror the U.S. female population, Sullivan notes. "I've talked to her about that . . . She shouldn't say it quite that way. With this kind of a sample, the wisest thing to do is talk about the percentages in the sample, and wonder whether one can generalize."

Some methodologists, researchers and statisticians who have examined Hite's approach go further, and say her women don't represent much of anything. Hite says she sent out 100,000 questionnaires; 4,500 were returned. "So few people responded, it's not representative of any group, except the odd group who agreed to respond," says Donald Rubin, chairman of the Department of Statistics at Harvard University.

"Isn't it a bit like the mass mailings that say, 'You've won $500, all you have to do is drive to Alaska and look at this land we have for sale'?" he asks. "I imagine you can get 5 percent of the people to respond to those ads, but I wouldn't want to say they're representative of the people you sent the letters to."

Rubin further points out that Hite herself has no assurance that even her claimed 4.5 percent response is correct. "How do we know how many people passed their hands over these questionnaires? You don't want to fill it out, you give it to your sister, she gives it to a friend. You'll get one response, but that questionnaire may have been turned down by five people. I didn't detect anywhere that she has any control over that fact."

"It may be even worse than that -- how do we know that one person didn't fill out 10 questionnaires? Maybe everyone's honest, but we know in other situations that people aren't above stuffing ballot boxes. And this is an emotionally charged issue."

Many of the essay questions are multipart and quite complex ("1. Who are you? What is your own description of yourself? . . . 88. What is sex with your partner (or in general) usually like? . . . 121. What things about women in general do you admire?"). Responses to these were converted by Hite into statistics. At the top of the sample questionnaire is the suggestion, in italics and with an exclamation mark: "It is not necessary to answer every question!"

Looking to the Experts

Corona Machemer, Hite's editor at Knopf, says, "I'm not prepared to comment on the accuracy of her research, or her method . . . Her research has been brought into question from day one, and has been defended from day one." Hite herself casually dismisses her critics, telling the Los Angeles Times that those who wonder about her research techniques "don't know methodology from a tree across the street."

Hite also said, in defending her statistics last Thursday on Oprah Winfrey's talk show, that "the presidents of the American Psychological Association, the American Sociological Association . . . have all said that my methodology is great."

"Good Lord," said Herbert Gans, ASA president and a sociology professor at Columbia University, when he heard of Hite's comment. He said he had received a copy of "Women and Love" from Hite the day the show aired, but hadn't read it.

"I don't know what it is. I know it's about sex, and I've heard the methodology is dubious," Gans said. Hite's comment was, he added, "unprofessional, not becoming of a social scientist, and a lie."

As for APA President Bonnie Strickland: "Scholars and scientists ask different questions and use different methods of inquiry. I think the book is very valuable, but as a professor of psychology I am not the best person to make a judgment about this particular type of methodology."

Hite has suggested that her critics haven't read her work, so they weren't making informed comments. So several methodology experts were asked by The Washington Post to look at "Women and Love." They raise serious questions about her data.

"If I'm going to talk about 'What is love?,' I can't just ask a handful of people," Hite said in an interview. "I have to ask a broad sample. To me, it's very important to measure different parts of the population, because there might be important differences."

Actually, the differences are minimal. Most women in Hite's study seem to feel the same way. Take one of her more controversial statistics: "70 percent of women married five years or more are having sex outside of their marriages." Here's the percentage breakdown by income:

Household income under $5,000: 71 percent are having affairs.

Income $6,000-$14,000: 72 percent.

Income $15,000-$39,000: 68 percent.

Income $40,000-$74,000: 69 percent.

Income over $75,000: 70 percent.

Poor, rich or middle class -- it makes no difference, all income levels are having affairs at nearly the same rate. The same is true no matter if a woman is white, black, Hispanic, Middle Eastern, Asian American or "other": the percentages range between 69 and 71 percent.

Nor does her age, level of education or employment status matter very much. Between 65 and 75 percent of the women are having affairs in the three age levels, three education levels, and four types of employment status.

This pattern is repeated throughout 99 pages of charts. The overwhelming majority of the women in Hite's study feel the same way about everything, and have had the same experiences. The numbers are nearly always close; sometimes they are exactly the same. Thirty-nine percent of women who have never been married have been struck or beaten by a lover; 39 percent of women married over 25 years have been struck or beaten by a husband or lover.

June Reinisch, director of Indiana University's Kinsey Institute, expresses skepticism over the similarity of the percentages. Look at Hite's divorce statistics, she says. "Can you imagine women who are making less than $5,000 expressing equal 'relief and well-being after a divorce' to ones making $75,000 or more? One can support her family, and the other is at the poverty level." (Hite's statistics are the same for each group: 70 percent.)

"Whether you're 18 or 71, you're going to answer the questions the same way; whether white, black, Hispanic, Middle Eastern or Asian American, you're going to answer the same way; whether you make under $5,000 a year or over $75,000, you'll answer the questions the same way," says Reinisch. "I've never seen anything like this in my career -- and the Kinsey Institute collects data from everybody."

Hite supporter Sullivan explains that this "could certainly happen given the nature of her sample. People sent these questionnaires back if they felt like it, and the people who sent them back had a particular motivation, and their motivation may have made them have more in common than their demographic differences would suggest."

Natural Selection

That explanation -- i.e., that the responses are the same because the women responding all felt the same, which is the only reason they responded in the first place -- doesn't satisfy William Simon, a sociologist at the University of Houston and coauthor of "Sexual Conduct: The Social Sources of Human Sexuality" and "The Playboy Report on the American Man."

"Let's say this is a self-selected sample -- women who have had lousy experiences with men are more likely to answer, and committed feminists are more likely to respond than nonfeminists. But a synonym for self-selected sample is a biased sample -- biased in favor of certain types of population."

"And if it's absolutely self-selecting, then the miracle is how this group could replicate with such precision the demographic profile of American women," says Simon, a former associate study director at the National Opinion Research Center and former senior research sociologist at the Kinsey Institute.

In a two-page table, Hite compares her study population with the U.S. female population. In two areas it differs: Her sample contains a smaller proportion of females under the age of 18 and a greater proportion of women with either some college education or college degrees.

But with income (which is subdivided into 10 groups); race/ethnicity (six groups); urban/rural (three groups); geographic region (four groups); marital status (six groups); labor force participation (six groups); and political affiliation (five groups), Hite is never more than 5 percentage points away from the national breakdown, and frequently is much closer. Thirty-three percent of the women in the country live in the South; 31 percent in Hite's study do. Or take an income example: 9.7 percent of the women in the country have incomes between $6,000 and $8,000; 10 percent of Hite's women do, too.

Gladys Engel Lang, a professor of political science and sociology at the University of Washington who wrote the "Methodological Observations" at the beginning of "Women and Love," says, "I'm not surprised {Hite's statistics match the U.S. female demography}, because she managed to get such a large number of respondents . . . I think she was very persistent on this."

However, Duane Alwin, a professor of sociology at the University of Michigan and a statistical methodologist at the university's highly regarded Institute for Social Research, says it's "very unusual that she could essentially match by haphazard methods the U.S. female population so closely. I don't see how she could have produced this close a match." (Alwin adds that "it's difficult to evaluate the quality" of Hite's work without getting better information on her methods and seeing her data.)

Martin Frankel, a professor of statistics and computer information systems at Baruch College, says: "If I could go and do a completely volunteer sample and get numbers that come back as close as hers without adjustment, I would consider myself the greatest sampling person in the world."

Says Judith Tanur, a professor of sociology at State University of New York, Stony Brook, and a specialist in survey methodology: "I find it difficult to believe that any amount of persistence in soliciting responses from targeted groups could ever match the U.S. population that closely in that many dimensions. The balancing act is just too hard."

Another problem, says Tanur: "When you get instructions to only answer those questions you wish to, you're likely to skip some. Isn't it more likely that, for example, a woman who feels strongly about affairs -- whether because she's having one or for some other reason -- would be more likely to answer questions on that subject, than a woman who does not feel as strongly?"

"Thus, her finding that 70 percent of all women married over five years are having affairs is likely to be wrong even for her study population. But because she does not report how many people answered each question, I cannot tell whether this means 70 percent of a thousand women, or 70 percent of 10 women, and so the statistic is meaningless."

"For similar reasons, I can't take any of her statistics seriously," Tanur says. "We know there is a segment of angry women out there. We have no idea how widespread this problem is."

Rubik's Cube

At her press conference, Hite told a questioner that, after getting back the first 1,500 questionnaires, "I did a breakdown of my sample and tried to see where I was lacking groups that would fit in." Then she targeted those specific groups -- in locations ranging from counseling centers to political groups to disabled people's organizations -- and sent them bundles of questionnaires. The 3,000 additional completed responses she got filled out her demographic balance. "I think that's extremely important and without that I would not consider my study valid," she said at the press conference.

But this process could be compared to Rubik's cube, Simon says: "As you go for one variable, you'd wind up distorting the others." If you need more women with household incomes under $2,000 to match the 18.3 percent in the U.S. female population, and if you somehow find out the groups these women belong to, what do you do if the ones who return the questionnaire are black, which pushes you over your limit in that category?

"What is the probability {that Hite could successfully do this} given the methods she used?" asks Hite supporter Sullivan. "It may be low, but it's still not zero."

Says Frankel: "Even in good samples, where you have a 50 or 70 percent response rate {instead of Hite's 4.5}, you usually see some skews -- say, with income, race or region. If she can do a sample like this, she's got the Rosetta Stone, and I'll come and study from her." Frankel has been chairman of both the American Statistical Association's section on survey research methods and the ASA's advisory committee to the U.S. Census.

There are, Frankel notes, ways to adjust the data to make your demographic distributions agree with predetermined distributions, such as the U.S. female population, "but this involves rather sophisticated weighting algorithms. And if she weighted the sample, then using the distribution to throw the patina of credibility around it -- which she does -- is terribly misleading."

No matter what, he adds, "the method of presentation she uses could lead both the survey expert and nonexpert to conclude that these figures are based on unweighted and unadjusted data."

In fact, editor Machemer and professors Engel and Sullivan say they don't think there has been any weighting or adjusting. "From what I understand, the statistics represent what she actually came up with," says Engel.

She adds: "I think her questionnaire was a very good one for what she was trying to study. I personally don't think it demonstrates that 80 percent of the people think this or 20 percent think that . . . Basically, I suppose if I were challenged on {the book}, I would say, 'Come in and look at the data. See if you would code it the same way I would code it.' " Engel herself hasn't seen the data. "I go by what she told me she did."

'Pardon the Phrase'

"I'm just trying to describe {contemporary relationships} as any scientist would -- pardon the phrase," Hite says.

But her results have been controversial, and "Good Morning America" decided to put an expert on with her. When Hite was informed of this, she objected.

"She didn't want to go on with anyone," says ABC spokeswoman Patty Neger. "She felt that when we had booked this originally, we booked it with her alone, and she wanted to keep it that way." The show gave in; Hite went on the air alone, although Neger says "GMA" toughened up its questions.

Another of Hite's many media appearances was supposed to be on the ABC TalkRadio Network's "Owen Spann Program." Spann says that after the first Hite Report came out, he was interviewing the author when "one of the callers asked her about her methodology -- how she arrives at her statistics. She became irate on the air, and basically refused to answer the question. We had a difficult time then."

This time, Spann said, he was "determined to be nice." So, right before air time, he complimented her on getting the Time cover. As Hite didn't like the story, she didn't like his comment. That started a three-way dispute among Spann, Hite and her husband. Spann asked Hite to leave, and she did.

In 1984, Hite sued writer Philip Nobile -- whose editorial in Forum magazine said she was one of the "charlatans" who needed to be driven "out of the erogenous zones" -- for making "false and defamatory statements." The $15 million suit was settled out of court, Hite's lawyer in the case says; neither party is allowed to disclose the terms of the deal. However, Nobile has apparently issued no public apology or retraction.

Yet for all the argument over Hite's numbers and what they really mean, for all the mystery surrounding Diana Gregory, even Hite's sharpest detractors feel the issues she's tackling are important ones.

"I do think there's something out there very much like what she's describing -- the deregulation of the family, how it's now a voluntary institution," says Houston sociologist Simon. "We're going through a transitional period -- new concepts, new norms, new frontiers."

"She's talking about something of fundamental importance, but to see it handled so superficially . . . It would be better if it were in the context of a much more thought-out product."


[ top ]