Infidelity

The scientific study of sexuality has faced the problem of finding a neutral terminology to describe behavior that often elicits strong moral sentiments. EMS, the abbreviation of extramarital sex, is a common convention used in scholarly papers. Philip Blumstein and Pepper Schwartz (1983) employ a nonjudgmental term, non-monogamy, which they apply to married and cohabiting couples, heterosexual or same-sex, who have sex outside of their union. Adultery is a narrower, legal concept. Adultery refers to voluntary sexual intercourse, either between a married man and someone who is not his wife or between a married woman and someone who is not her husband. Although an unmarried partner may have an adulterous relationship with a married person, only married people have extramarital sex. The epidemiological literature in public health focuses on the number of sexual partners. This approach to measuring sexual behavior distinguishes secondary sex partners, who are defined by reference to a primary sex partner (i.e., the person reported to be the most important or frequent sexual partner). The primary partner is typically a spouse, cohabitor, or steady date. Married people who have multiple or secondary sex partners are assumed to have had extramarital sex.

The accuracy of sex data depends on the respondents' recall and candor. People have difficulty remembering sexual activities from the distant past. Also, because sexual infidelity is what survey experts describe as "sensitive" behavior, people may be embarrassed or reluctant to admit infidelities, particularly if an infidelity is not really characteristic of their usual patterns. This reporting bias could mean that survey estimates of extramarital sex and secondary partners are understated. Critics of sex surveys have challenged the validity of data, because men, on average, report a higher number of partners than women do. This pattern is seen in the United States, Great Britain, Norway, and Canada. Close examination of sex data does not suggest widespread problems. The problem is limited to a few men who skew the results by reporting extremely high numbers of sex partners.

Data quality is not a new concern. To discourage under-reporting of sexual behavior, Alfred Kinsey's pioneering sex studies of the 1930s and 1940s used complex cross-checks and aggressive interviewing techniques. Kinsey's estimates of the population engaging in extramarital sex—half of married men at some point and a quarter of married white women by age 40—were startlingly high. As statistical experts of the day noted, it is impossible to determine the validity and reliability of Kinsey's findings. His figures may have resulted from his biased volunteer sample, which was skewed toward prisoners, divorcees, and others whose sexual experiences were not representative of the U.S. population at large. The limitations of the historical data make it impossible to determine with much confidence whether the incidence of sexual fidelity has changed over time for U.S. husbands and wives.

Because of the sensitive nature of their topics, sex studies, including recent ones, have encountered heated political opposition. As a consequence, much research on extramarital sex has been based on dubious sources, such as readers who are sufficiently motivated to mail back a magazine questionnaire on sex. Largely in response to the AIDS crisis, however, several countries fielded large, nationally representative sample surveys of sexual behavior in the 1990s. In the English-speaking world, two surveys in 1992—the British National Survey of Sexual Attitudes and Lifestyles (Wellings et al. 1994) and the U.S. National Survey of Health and Social Life (Laumann et al. 1994)— have contributed to our understanding of sexual partnering.

Contemporary interview surveys have devoted considerable attention to improving the quality of sex data. Researchers go to great lengths to develop, pretest, and refine their questionnaires. To insure the integrity of their scientific samples, they work hard to secure interviews with sample persons who are difficult to locate or reluctant to be interviewed. They make special efforts to conduct confidential interviews out of earshot of other household members. Anonymous, self-administered questionnaires that work well for sensitive questions are combined with face-to-face interviews where clarification is needed. For example, interviewers collect complicated rosters for the start and end dates of sexual relationships; these can be used to determine if there are overlaps in time that would indicate sexual infidelity. Data are analyzed for consistency and compared to results from other surveys.