Psychological Perspectives on Human Sexuality / Edition 1 available in Hardcover
Psychological Perspectives on Human Sexuality / Edition 1
- ISBN-10:
- 0471244058
- ISBN-13:
- 9780471244059
- Pub. Date:
- 12/23/1999
- Publisher:
- Wiley
Psychological Perspectives on Human Sexuality / Edition 1
Hardcover
Buy New
$194.75Buy Used
$98.68-
-
SHIP THIS ITEM
Temporarily Out of Stock Online
Please check back later for updated availability.
-
Overview
Product Details
| ISBN-13: | 9780471244059 |
|---|---|
| Publisher: | Wiley |
| Publication date: | 12/23/1999 |
| Pages: | 696 |
| Product dimensions: | 7.24(w) x 10.20(h) x 2.05(d) |
About the Author
Read an Excerpt
Chapter 1
Research in Human Sexuality
David A. Wagstaff, Paul R. Abramson, and Steven D. Pinkerton
In this introductory chapter, we provide an overview of many of the important elements that define contemporary research on human sexuality. We hope that our material facilitates the reader's enjoyment and appreciation of the interesting and informative chapters that follow. Given the limits of time and space, we had to be selective and undoubtedly excluded a number of important research areas and developments. However, we have tried to provide a broad sampling of the many flavors that constitute contemporary sex research. In this chapter, we touch on such fundamental questions as What (and who) do sex researchers study? What kinds of issues do sex researchers examine? What methods do they use?
The chapter is loosely organized in four sections. In the first, we consider definitions and, specifically, the meanings that individuals and researchers associate with the words "sex" and "research."
In the second section, we discuss some of the theories that guide sex research programs and currently provide the theoretical basis for interventions that are designed to prevent sex-related social problems (such as sexually transmitted infections and unwanted pregnancies).In the third section, we consider some of the methods used to collect data on human sexuality. We close, in the last section, with a discussion of clinical applications and our perspective on the future of sex research.
Preparation of this chapter was supported, in part, by NIDA center grant P50-DA10075 awarded to the Center for the Study of Prevention through Innovative Methodology, Pennsylvania State University, and by NIMH center grant P30-MH52776 awarded to the Center for AIDS Intervention Research, Medical College of Wisconsin
.WHAT IS SEX?
Sex, like love, is a many splendored thing. It also has many definitions, which have given rise to an equal number of misconceptions. One of the most fundamental confusions concerns the difference between "sex" and "gender." Sex, of course, has a dual meaning, referring both to a physical activity and to a physical characteristic. The latter is often confused with gender (roughly, one's sense of femaleness or maleness), which is perceived internally and negotiated (i. e., socially constructed and affirmed through interaction with other individuals). The distinction is subtle. Although most sexual characteristics form a continuum from "male" to "female," the genome of most males has both an X and a Y chromosome, and most men exhibit typically male secondary sexual characteristics, such as penile development. The male sex is thus defined by physical characteristics. In contrast, the male gender is defined by a confluence of psychological and social considerations. At the psychological level, we can define a person's gender identity as his or her subjective self-perception of maleness or femaleness. On this basis, gender identification is inherently psychological: a physiologically male individual can perceive himself as female, and vice versa. However, cultural influences are also evident in gender identification. Culture defines what it means to be male or female to the extent that the individual's culture determines the appropriate gender roles for men and women. Indeed, the term "gender" is often used to denote the cultural, social, and psychological experience of belonging to a particular sex or fitting into a particular gender role. For example, one might ask, What is it like to be a woman in contemporary America?
Sex researchers concern themselves with all the various meanings of "sex" and "gender," as well as a number of related and not-so-related issues. The long list of topics they study includes ( but is not limited to) the physiology and anatomy of the reproductive tract; potential physical determinants of maleness and femaleness, including genetics, hormones, and neurophysiology; the intricacies of human sexual responses, including mechanisms of arousal, orgasm, and resolution; sexual attraction, whether heterosexual, homosexual, bisexual, asexual, or some mix; individualistic expression of sexuality and what it means to adopt, or identify with, a particular sexual orientation; the role of fantasy in healthy and "deviant" sexuality; the etiology, prevalence, and expression of various paraphilias (nonnormative sexual attractions and practices), such as necrophilia, bestiality, sadomasochism, and various fetishes; what it means to be male or female in a particular culture, and the determinants of gender identity; the social and individual ramifications of sexual behavior, including unwanted pregnancies and sexually transmitted infections (STIs) such as human immunodeficiency virus (HIV); the commercialization of sex, especially pornography and prostitution; and the psychological determinants and consequences of sexual behavior and ideation.
Some sex researchers specialize in tracking the history of sexual mores; in how societies regulate sexuality; in how human sexual behaviors compare with those of closely related primates; or in cross-cultural comparisons of sexual expression. Given the importance of sex to the continuation of the human species, it is perhaps not surprising that the study of human sexual behavior should encompass such an extensive and varied range of topics. (Like the rest of this volume, however, the remainder of this chapter will focus primarily on psychological aspects of human sexuality.)
As the preceding discussion suggests, the vocabulary of sex research contains some ill-defined terms. However, the use of some inexact language in science is not uncommon. As Peter Medawar observed, biology would not exist if its technical terms had to be defined precisely (Medawar & Medawar, 1983). For different reasons, the same may be true for research in human sexuality. This ambiguity in the meanings assigned to and the uses of sexual terminology arise from two inextricably linked sources: intrapsychic variability and cultural differences. The former refers to the idiosyncratic ways that people define sex and sexuality for themselves; the latter refers to the ways that different cultures define these concepts.
INTRAPSYCHIC DIFFERENCES
There are many interesting anecdotes that illustrate the ways that different individuals think about sex. Our favorite is the following. A psychology professor was studying contraceptive neglect among pregnant teenagers attending a support group for unwed mothers. Prior to assessing their contraceptive experiences, the professor asked a number of standard sex questions, including, "Are you now or have you ever been sexually active?" One teen answered no. Puzzled by this enigmatic response, the professor asked the pregnant teen how she could claim she was never sexually active. Her telling reply was, "I just kind of lie there." To this teen, the phrase "sexually active" referred to the amount of energy that one put into sex! The professor and every other study participant understood the phrase differently. To them, it meant engaging in sexual intercourse-- the active participation of all parties was not required.
The phrase "sex differences" also means different things to different people. When some individuals use the phrase, they are often referring to gender, gender roles, and/or gender differences. The latter are critical to sexuality itself, both on the psychological level and the genetic level (Margolis & Sagan, 1986). Recently, the concept of two genders has been challenged with the view that there are (or can be) more than just male and female (L. Cohen, 1995; Herdt, 1994). The Hijra of India provide a fascinating example (L. Cohen, 1995). The Hijra are castrated genetic males, who perceive themselves as neither male nor female, and whose social and sexual roles do not fit into either category. Thus, they are a "third gender," a group of individuals who do not fit neatly within the traditional strictures of sexual dimorphism. These examples are meant to emphasize the fact that researchers need to choose their words carefully. They also suggest that the way individuals think about sex is curiously personal-- and the way individuals think about sex often influences or reflects how they act. For example, individuals who believe that masturbation is sinful express and experience the act differently from those who believe that it is essential to sexual health (Abramson & Mosher, 1975, 1979; D. Mosher & Abramson, 1977; Pinkerton & Abramson, 1999).
Accommodating these kinds of individual differences, particularly in a country as culturally diverse as the United States, is a difficult task. In large national surveys of sexual behavior, a standard terminology is clearly needed; therefore, clinical language is typically used (Binson & Catania, 1998). However, not all research participants understand the clinically correct terms for the sexual anatomy and for common (and uncommon) sexual behaviors. Indeed, as many as 25% of Americans with fewer than 12 years of schooling may have difficulty understanding terms such as "vaginal intercourse" and "anal intercourse," which frequently appear in the instruments that researchers use to obtain data on sexual behaviors that confer risk for HIV transmission (Binson & Catania, 1998). In addition, the evidence suggests that men are more likely than women, and minority respondents are more likely than White respondents, to have trouble understanding these terms. These findings suggest that, whenever possible, survey respondents should be provided with definitions of the relevant terms (e.g., "vaginal intercourse") before being asked questions about their own behavior. Additionally, these findings suggest that researchers should ensure that the questionnaires and interview forms used to obtain sexual behavior data are tailored to the population of interest and are both culturally and developmentally appropriate. In some cases, researchers will have more success if they use slang or "street" language to describe sexual behaviors and show greater sensitivity when asking questions (Mays & Cochran, 1990).
Ultimately, sex researchers need to be clear about the definitions they use and acknowledge the fact that other individuals-- especially research participants, but also other researchers-- may use different definitions. Unfortunately, there is no "one size fits all" solution to this difficulty. Although some individuals may feel more comfortable discussing their sexual behavior using clinical terms such as "vaginal intercourse" or "cunnilingus," others may not understand these terms, or may instead prefer to use slang or vernacular equivalents. The challenge for sex researchers is to correctly gauge which terms would be most acceptable and appropriate given the particular research population and the aims of the study. Conducting focus groups or otherwise eliciting input from the study population prior to finalizing survey or interview instruments can be extremely helpful in this regard.
CULTURAL DIFFERENCES
In the past, people believed that the cultural world followed the same laws as the natural world. In particular, the cultural world was viewed as an adaptation designed to best serve the needs of men and women. As such, they believed in an objective cultural reality that could be reliably ascertained: those who studied the cultural world merely needed to observe it and deduce the functions it served. This was as true for sex as for any other phenomenon worthy of investigation.
This view changed in large part as a result of the work of postmodern writers such as Foucault (1978) and Sedgwick (1990). These writers "deconstructed" sex, demonstrating that the concept of "natural" sex was highly capricious and often politically or religiously motivated. In their view, although heterosexuality might be prized by most societies, there is no inherent validity for this valuation. Instead, the greater value placed on heterosexuality arose from pervasive religious emphases on procreation and the (heterosexual) relationships that promote this end. Thus, if heterosexuality were considered more "natural" than homosexuality, it was only because society deemed it so.
Thoughts on the nature of female orgasm provide another example of the social construction of sex. In the sixteenth century, particularly in midwifery manuals, female orgasm was presumed routine and necessary for conception (Laqueur, 1990). (Some formulations of this belief held that women, like men, released "sperm" when they had an orgasm, and that it was the mixing of the male and female sperm that formed a fetus.) However, by the nineteenth century, many doctors had rejected this popular theory and began to doubt the reality of female orgasm (i.e., they doubted that women were capable of orgasm). The two views reflect radically different conceptualizations; if holding one of them made it easier or more difficult for a woman to achieve orgasm, we should not be surprised. Psychologists have long known that "expectancy is a self-fulfilling prophecy."
Both cultural variability and the cultural construction of sex are evident in historical and contemporary conceptualizations of homosexuality. Japan has a long and celebrated history of male homosexuality, particularly among the samurai and certain sects of Buddhist monks (Leupp, 1995; Pinkerton & Abramson, 1997). There are no legal proscriptions on homosexuality or laws concerning the practice of sodomy in Japan. This historical experience contrasts sharply with that of other countries, particularly the United States, where homosexuality has been vigorously penalized and pathologized. Indeed, the Western concept of homosexuality-- in which homosexual behavior is considered deviant and the practice thereof defines someone as a different sort of person (i. e., a "homosexual")-- is basically a nineteenth-century invention. It differs significantly from the way other cultures organize erotic desire for persons of the same sex. In the Sambia culture of Papua, New Guinea, for example, all teenage boys are expected to undergo a period of ritual homosexual behavior prior to establishing a lifelong, heterosexual marital relationship. The Sambia do not consider the boys who engage in oral sex with one another to be "homosexual," but instead believe that the ingestion of semen plays a critical role in masculinizing the receptive partner, helping him to achieve the strength and courage required of men in this fierce warrior society. In contrast, in the traditional Mexican culture of male "machismo," the "passive" or receptive partner of a male-male sexual dyad is considered to be homosexual, whereas the "active" or insertive partner can still claim a heterosexual identity (Carrier, 1995). Finally, in the United States, Kinsey's claim that approximately 10% of the male population had engaged in at least one same-sex encounter has been widely misinterpreted to mean that 10% of the male population is homosexual.
Thus, when viewed cross-culturally or historically, it is clear that there is no standard set of behaviors or feelings that constitute the construct of homosexuality (Greenberg, 1988; Herdt, 1997). Similar comments apply to many other fundamental conceptual constructs.Such examples have clear implications for research on human sexuality. First, they suggest that many sex-related constructs-- the bread and butter of the psychological study of human sexuality-- are socially constructed, rather than fixed aspects of the natural world. They also demonstrate the immense variability of these constructs when viewed historically or cross-culturally. Third, they remind us that the meaning given to a construct by a culture is often politically or religiously motivated. Thus, exceptional care should be taken in how such constructs are defined and investigated. Finally, as noted in the preceding section, the instruments (e.g., questionnaires and interview forms) used to measure individual sexual behavior should be culturally appropriate for the target population.
A PHILOSOPHY OF SCIENCE
To conduct research, it is essential to understand the purpose and methods of research, as well as the philosophical foundations of the scientific enterprise. In this section, we briefly summarize the contributions of three important philosophers of science: Francis Bacon, Thomas Kuhn, and Sir Karl Popper.
MODELS OF THE SCIENTIFIC ENTERPRISE
In the prescientific era, people believed they could determine if a statement was true by consulting religious texts or previous work by respected authorities (preferably Greek). Francis Bacon (1620/ 1956) changed this when he proposed that researchers use observation and experimentation to determine the validity of scientific conjectures. In particular, he proposed that researchers observe and make records of the phenomenon in which they are interested, and then develop a theory on the basis of what they observe, a process known as scientific induction. Theory was important because it facilitated interpretation of the available data and yielded predictions that the researcher could test through further experimentation. If the findings agreed with the researchers' predictions, the theory survived; if they did not, the theory was revised to take into account the new observations. Thus, research was conceived as an iterative, self-correcting process, in which theory and observation are partnered with an experimental method. According to the Baconian model, "normal science" operates via a three-stage process of hypothesis formation, testing, and revision. Scientific knowledge grows as more elaborate and extensive theories are proposed and tested according to the above model. Thomas Kuhn (1962) challenged this view of normal science and offered an alternative explanation for the growth of scientific knowledge and the origin of scientific revolutions. Scientific revolutions are reflected in conceptual changes of a fundamental kind and radical alterations in the standard or accepted explanations for natural phenomena (J. Cohen, 1985). The emergence and acceptance of Darwin's theory of natural selection is an obvious example. Kuhn added to our understanding of scientific revolutions by highlighting the manner in which such revolutions evolve. Contrary to prevailing wisdom, Kuhn argued, scientific revolutions do not come about as the result of the practice of normal science, or the accumulation of evidence from successive experiments, in accordance with the Baconian method. Instead, Kuhn believed that scientific revolutions result from so-called paradigm shifts, precipitated by a crisis in the current state of science. That is, with the emergence of evidential patterns that cannot be explained by existing theories, a crisis occurs, followed by a revolution that produces a new paradigm. The theory of natural selection illustrates Kuhn's argument. This theory did not result from a series of prior experiments, but from Darwin's careful observations of nature. In time, natural selection was accepted because it provided a more robust theory than the prevailing religious and Lamarckian explanations of evolutionary change.
Sir Karl Popper's (1972, 1983) work has been highly influential and represents another radical departure from normal science. Popper (1983) contends that a theory should be accepted as a statement of empirical science (as opposed to a statement of belief, for example) if and only if it is possible for researchers to collect data that could potentially refute it. Like Bacon, Popper makes the clear distinction between a scientific theory that can be proven false and religious dogma that cannot. He believes that "scientific theories are distinguished from myths merely in being criticizable, and in being open to modification in light of criticism" (1983, p. 7). In fact, Popper concludes that the unifying characteristic of all "true" sciences is that their theories are subject to criticism and revision. Researchers must be able to conduct independent tests of each leading explanation of the phenomenon of interest; the more rigorous and frequent the test, the more satisfactory the explanation (Popper, 1957, 1972). Scientific progress, then, is marked by better and better explanations (Popper, 1972).
In summary, theory and experimentation form the basis of scientific research. In our view, the role of experimentation is multifaceted. It can be used to verify a hypothesis or to discredit it in accordance with Popper's doctrine of falsifiability, or it can be used to more thoroughly investigate a phenomenon and thereby assist with the formulation of new or better explanations.
PRINCIPLES OF STUDY DESIGN
Although the scientific method as outlined in this chapter provides a framework for the conduct of certain types of research, it is not often applied to the study of human sexuality, which relies instead on other modes of inquiry and validation. Residing primarily in the social sciences, the study of human sexuality has embraced methods characteristic of the early life (agricultural) and social sciences, particularly in the design of experiments and the analysis of data.
As W. Cochran (1976) suggests, the seeds of contemporary study design and data analysis were sown in English fields in the eighteenth and nineteenth centuries. The best of these comparative agricultural experiments used systematic layouts, careful measurement, replication, and concurrent controls. A major advance in statistical practice occurred in 1908 when William Gosset, writing under the pseudonym of Student, published a paper on the distribution of errors about the sample mean (Boland, 1984; G. Box, 1984; J. Box, 1981; Student, 1908). The statistic and distribution were subsequently named in Student's honor (i. e., Student's t-statistic). Gosset's paper was important because it showed that it was possible to study the exact distribution of a sample statistic. In addition, it showed how one could use data to draw conclusions when the sample size was small. Prior to the paper's publication, statistical theory was based primarily on what was known about the distribution of statistics in large samples (and, in fact, many years would pass before researchers used the small sample statistic that Gosset proposed in his paper).
Extensive use of probability theory in the design and analysis of agriculture experiments began in 1919 when R. A. Fisher joined the Rothamsted Experimental Station as its first statistician (G. Box, 1984; J. Box, 1980). Fisher was interested in determining how experiments might be designed so that they provide the clearest answer. In the first article that he published on the subject, R. Fisher (1926) introduced the principle of randomization and the use of factorial designs, which required the researcher to vary simultaneously two or more factors that were believed to affect the outcome. (In a text published in the preceding year, R. Fisher (1925) introduced a larger audience to the analysis of variance, tests of statistical significance, and the 5% significance level.) Unlike earlier researchers who were concerned with obtaining a precise estimate, Fisher was concerned with a study's validity and efficiency, recognizing that statistical rigor could not substitute for inadequate design. Fisher's proposed use of randomization (into treatment and control groups) was not accepted initially by his colleagues, who favored the use of systematic designs. However, his arguments eventually gained broad acceptance, and randomization became the distinguishing characteristic of the "true" experiment and the standard against which most social science research studies, including studies of human sexuality, are judged today. Of no less importance to the current study of human sexuality were the other principles that Fisher introduced or championed: factorial arrangements (and the idea that outcomes are determined by multiple causes), replication, the use of concurrent controls, and careful measurement (a principle that is reaffirmed when researchers assess the reliability and validity of their scales and instruments).
THEORY
In the physical sciences, theory development represents a search for the laws of nature. Such laws are presumed to exist; however, it has proven difficult to find laws that hold without exception, exactly, and throughout time (Kemeny, 1959). Where sex is concerned, the goal is often more humble. What many researchers want from a theory of human sexuality is a good explanation. Thus far, sex researchers have had to be satisfied with good explanations because research findings have yielded relatively few consistently verifiable facts. The latter are required to construct a scientific theory and, in particular, a fully mathematized theory capable of expressing universal laws of behavior (Abramson, 1990).
CONTRIBUTIONS FROM THE LIFE SCIENCES
Darwin's theory of evolution by natural selection represents one of the few exceptions. It is a "true" scientific theory that appears to be a law of nature. Evolutionary theory has also given rise to a fertile area of sexuality research known as evolutionary psychology by some, and as sociobiology by others. This research focuses on how the physical and social environments in which humankind evolved have shaped the way we think and act. The fundamental theoretical assumption that underlies this incipient program of research is that many (if not most) physical and behavioral traits evolved millions of years ago as solutions to problems of relevance to either survival or reproduction (these behavioral solutions are known as "adaptations"). Evolutionary psychology differs from other evolutionary theories, including sociobiology, in its explicit focus on the innate psychological mechanisms through which particular adaptations are expressed. (For more on evolutionary psychology, see Buss, 1994; Cosmides & Tooby, 1987; H. Fisher, 1982, 1992; Symons, 1979, 1992, 1995. For additional information on sociobiology, see Barkow, 1980; E. Wilson, 1975.)
According to evolutionary psychological theory, the physical features that people find attractive are precisely those that have been reliably correlated with reproductive success throughout human existence. Symons (1995) argues that the appearance of youth and healthfulness is universally associated with female sexual attractiveness because these qualities act as de facto markers for reproductive fitness. Consequently, he predicts that in all societies in which women attempt to alter their appearance-- through cosmetics, diet, exercise, and dress-- they will do so in a manner that accentuates the appearance of youth and health. Conversely, men are assumed to have inborn mechanisms for detecting these qualities in women (Symons, 1995). Notice that in this example, Symons begins with a broad theoretical framework (evolutionary psychology), which is then narrowed down to a specific, testable hypothesis (women can be expected to accentuate the appearance of youth and health).
There are several notable biologically based theories of human sexual behavior and sexual desire. Most biologically based theory and research owes a debt of gratitude to the pioneering work of John Money and his colleagues. Money argued for developmentally focused research on sexuality (including sexual orientation, gender differentiation, and gender identification). In this model, sexuality begins in the womb, under the combined influences of genetics and hormones (Money & Ehrhardt, 1972). Recently, Dean Hamer and his colleagues gained much notoriety for identifying a genetic site that appeared to be correlated with an increased incidence of familial male homosexuality (Hamer, Hu, Magnuson, Hu, & Pattatucci, 1993). Genetic influences are also apparent in studies of twins: even when raised separately, twins are more likely to have the same sexual orientation than are non-twin siblings (because twins share more genes than do non-twins, this suggests that sexual orientation may have a heritable component) (e.g., Bailey & Pillard, 1991; Bailey, Pillard, Neale, & Agyei, 1993). This does not, of course, explain how genetic differences influence sexual behavior and especially sexual orientation. One possibility, popularized by Simon LeVay (1991, 1993), is that genetic differences lead to neuroanatomical and/or neurophysiological differences between gay and straight individuals.
CONTRIBUTIONS FROM THE HUMANITIES
The field of human sexuality is also deeply indebted to theories developed within the humanities, in particular, those founded on postmodernism, deconstructionism, and feminism. As mentioned previously, the works of Michel Foucault and Eve Kosofsky Sedgwick have been highly influential, as have the writings of Thomas Laqueur (1990) and Judith Butler (1990). Collectively, these authors have provided insights on the ephemeral nature of sexual constructs. Their theories have been elaborated upon and extended by many authors. To name a few: Marjorie Garber (1992) has considered cross-dressing; Lillian Faderman (1991), the history of lesbian life; Constance Penley (1997), sex in popular culture (notably, "slash fiction," in which fictional characters, such as Star Trek's Spock and Kirk, are cast into improbable sexual relationships); Martine Rothblatt (1995), the historical and legal stature of binary gender differentiation; Cindy Patton (1985), sex and its relation to disease; Camille Paglia (1990), art and decadence; Laura Kipnis (1996), pornography; Carol Vance (1984), pleasure and danger; and Jeffrey Weeks ( J. Weeks & Holland, 1996), sexual civilizations.
CONTRIBUTIONS FROM THE COGNITIVE SCIENCES
In another area of active theorizing, several researchers have proposed models to explain how women and men behave in sexual situations. One of the best known of these is Simon and Gagnon's "script theory" (1984; Gagnon & Simon, 1973), which borrows from the more general notion of behavioral scripts that has been popularized by cognitive scientists (e.g., Schank & Abelson, 1977). As the name implies, a script is a stereotypical set of behavioral responses and expectations that are appropriate for a particular scenario. For example, a restaurant script might include expectations regarding the waiter's behavior (" He will bring me a menu"), the service that will be provided (" They will cook and serve me food"), and the need to pay for what is ordered, as well as particular behavioral responses that will be required (taking a seat, choosing from the menu, ordering, eating, tipping, paying, etc.). Because the interaction follows a well-defined script, the individual knows what to expect and how to act. According to sexual script theorists, people also have scripts for how to behave in sexual situations, including specific scenarios for dating, foreplay/sexual play, and intercourse. Recently, script theory has been applied to the question of how men and women choose when (and with whom) to practice safe or unsafe sex (Seal, Wagner, & Ehrhardt, 1999).
CONTRIBUTIONS FROM MATHEMATICS
Although a number of mathematical models of sexual behavior have recently been proposed or reexamined within the context of the ongoing HIV epidemic, this area remains one of the least well-developed areas of sexual theorizing. Most existing models focus on the relationship between sexual behavior and viral transmission, but tell us little about human sexual behavior per se (E. Kaplan, 1995). Among the most influential sexual transmission models are the population-level models of Anderson and May (1988; May & Anderson, 1987); the individual-level, Bernoulli-process model examined by Pinkerton and Abramson (1993, 1998); and the social network models of Morris (1994; Morris & Kretzschmar, 1995; Morris, Pramualratana, Podhisita, & Wawer, 1995; Morris, Zavisca, & Dean, 1995).
CONTRIBUTIONS FROM PREVENTION RESEARCH
Health psychologists and prevention researchers have used a number of theories to design programs and behavior change/risk-reduction interventions that address adolescent pregnancy and the transmission of sexually transmitted infections (especially HIV). Four are described here: the Health Belief Model, the Theory of Reasoned Action, Social Learning/Self-Efficacy Theory, and the Theory of Transtheoretical Change.
The Health Belief Model has been the focus of numerous studies (Becker, 1974; Brown, DiClemente, & Reynolds, 1991; Condelli, 1986; Eisen, Zellman, & McAlister, 1992; Hiltabiddle, 1996; Janz & Becker, 1984; Kirscht & Joseph, 1989; Maiman & Becker, 1974; Rosenstock, Strecher, & Becker, 1988, 1994). The model assumes that an individual's choices and subsequent behaviors reflect a rational decision-making process. Further, it assumes that an individual's decisions are based on her perception of how susceptible she is to the threat (e.g., unwanted teenage pregnancy), how severe the consequences are, and the relative costs and benefits of adopting the various risk-reduction behaviors (use an effective contraceptive, remain abstinent).
The Theory of Reasoned Action (Fishbein & Ajzen, 1975) and the Theory of Planned Behavior (Ajzen, 1991; Ajzen & Madden, 1986) are similar to the Health Belief Model: all three are cognitive models and, as such, do not give much weight to emotion or to individuals' motivations and drives. The Theory of Reasoned Action focuses on individuals' intentions to engage in the target behavior. For example, an individual's intentions to use condoms are assumed to be a function of his or her attitudes toward condom use, and the salience and valence of the condom use attitudes held by significant others (friends, sexual partners, parents). Despite the fact that it provides no direct role for emotion, the Theory of Reasoned Action has been applied extensively to the study of sexual behavior (Baker, Morrison, Carter, & Verdon, 1996; S. Cochran, Mays, Ciarletta, Caruso, & Mallon, 1992; Fishbein & Middlestadt, 1989; W. Fisher, Fisher, & Rye, 1995; Hecker & Ajzen, 1983; L. Jemmott & Jemmott, 1991; Morrison, Gillmore, & Baker, 1995; Terry, Gallois, & McCamish, 1993).
Bandura's Social Learning Theory (1977b) and his Self-Efficacy Theory (Bandura, 1977a, 1982; Strecher, DeVellis, Becker, & Rosenstock, 1986) are also widely used in behavior change and HIV prevention research. The former reminds researchers that sexual behaviors, like most behaviors, are learned. The latter serves to focus attention on the fact that individuals will adopt a recommended behavior (e.g., use condoms with a new partner) only if (a) they believe that they can enact the behavior (ask and, if necessary, insist that the new partner use condoms) and, (b) they believe that the behavior will achieve the desired outcome (prevent HIV infection and/or pregnancy). Because the related social cognitive theory is used frequently in clinical practice, Social Learning Theory and Self-Efficacy Theory have provided the rationale for many sexual risk-reduction interventions (Bandura, 1989; Basen-Engquist & Parcel, 1992; Heinrich, 1993; J. Jemmott, Jemmott, Spears, Hewitt, & Crus-Collins, 1992; Joffe & Radius, 1993; Kasen, Vaughan, & Walter, 1992; L. Lawrance, Levy, & Rubinson, 1990; Levinson, 1986; D. Rosenthal, Moore, & Flynn, 1991; Schinke, Holden, & Moncher, 1989; K. Weeks, Levy, Zhu, et al., 1995).
The fourth theory that we mention is the Theory of Transtheoretical Change (Prochaska & DiClemente, 1983). This theory differs from the three other theories of behavior change in that it views change as a process, specifically, a sequence of five stages: precontemplation, contemplation, preparation, action, and maintenance. Although it has its origins in studies of smoking cessation, it has been applied to sexual behavior change and in the design of STI/ HIV prevention interventions (Centers for Disease Control and Prevention, 1993; Grimley, Prochaska, Velicer, & Prochaska, 1995; Grimley, Riley, Bellis, & Prochaska, 1993; Santelli, Kouzis, Hoover, Polacsek, Burnell, & Celentano, 1996).
INTEGRATIVE THEORIES OF HUMAN SEXUAL BEHAVIOR
There are surprisingly few overarching theoretical integrations of human sexual behavior. As a consequence, fundamental questions such as Why does sex feel good? and What determines erotic preferences? have largely been left unexplored. We note two exceptions to our generalization.
In a recent book, Abramson and Pinkerton (1995) examined the central role of pleasure in motivating and shaping human sexual experience. They argued that reproduction is a by-product of the human desire for sexual pleasure, rather than the other way around. The implications of this seemingly obvious but ultimately nontrivial shift in how individuals conceptualize sexuality extend from the debate over the legitimacy of homosexuality, to the prosecution of pornographers, to the practice of safer sex.
The work of Robert Stoller provides another example of a theoretically rich investigation of human sexuality, and especially "deviant" sexuality (e.g., see Stoller, 1985; Stoller & Herdt, 1985; Stoller & Levine, 1993). Stoller, a psychoanalytic psychiatrist by training, hoped to answer basic questions about human sexual desire by exploring the "fringes" of socially normative behavior. For instance, he offers the following psychoanalytic explanation for the development of male transvestism (Stoller, 1971). According to Stoller, as young boys, most transvestites suffer the humiliation of being dressed in girl's clothes by their (usually female) relatives-- a symbolic act of castration. This leads to later cross-dressing because, for the adult transvestite, dressing as a woman reaffirms his masculinity by juxtaposing the threat of castration symbolized by his feminine attire with that supreme marker of maleness, his intact penis.
Although our task is to consider research on human sexuality, we recognize that there is much to be gained by examining theories of primate sexuality. Here, we recommend the works of Mary Pavelka (1995), Frans de Waal (1989, 1995), and Kim Wallen (1995).
METHODS
This section discusses several important methodological considerations for conducting human sexuality research. The first issue concerns how individuals are selected for participation in a study, or more precisely, how individuals are sampled from a population, with the hope that the results obtained from this sample can be generalized to the larger population. Next, we discuss some of the many methods that have been developed for obtaining information about people's sexual behavior, including surveys, interviews, focus groups, direct observation, and clinical and laboratory research.
SAMPLING
Sampling has come to play a critical role in both qualitative (descriptive) and quantitative research on human sexuality. Because researchers do not have the time, money, ability, or inclination to observe all members of the study population, they must observe and assess a select number of individuals from the population (the study sample). Sampling theory, methods, and practices ensure that an adequate number is selected and that the individuals selected are representative of the larger population. They also serve as a form of social control, ensuring that researchers have not purposefully selected individuals with characteristics that favor the researcher's hypothesis.
One example of a conventional sampling design is provided by the school-based surveys that assess sexual behaviors among high school students (Centers for Disease Control and Prevention, 1990). With a conventional design, selection of the classrooms and of individuals within each classroom does not depend on what has been observed in other classrooms or in the particular classroom. Once the sampling frame (roughly, a list of the schools, classrooms, and students to be considered for inclusion in the study) is determined, the probability that an individual will be selected as a member of the study sample can be calculated. Imagine that a researcher is interested in determining the proportion of sexually active high school students who drink beer or other alcoholic beverages shortly before intercourse. If the sampling design calls for the researcher to sample schools, then classrooms within schools, and then individuals within classrooms, each individual will have a known probability of being included in the study sample. The inclusion of John, who has sex with Jennifer and is a close friend of Joe, will not affect Jennifer's chances of being included (and vice versa). More important, if only a few students within the selected schools have had sexual intercourse, there may be an insufficient number to establish the prevalence and determinants of the behavior of interest (alcohol use before sex). When the researcher does not take into account the distribution of the characteristic in the population, a conventional sampling design will probably provide a poor estimate of the characteristic that the researcher wanted to assess (Thompson, 1997).
Compared to conventional sampling designs, ethnographic studies have a much greater potential to reach a larger number of individuals who possess the characteristic or engage in the behavior. Such studies use a variety of techniques such as snowball sampling or chain referral sampling. Snowball sampling may be the best known of the various techniques (Biernacki & Waldorf, 1981; Faugier & Sargeant, 1997; Van Meter, 1990). The term has been used to describe two types of network sampling procedures. For one type, a few identified members of the population are asked to identify other members who are asked to identify other members of the population (Kalton & Anderson, 1986). This procedure is useful for generating a sampling frame. For the second type of snowball sampling, members are asked to identify a fixed number of other members, who are asked to identify other members, for a fixed number of rounds (Goodman, 1961). This procedure is useful for identifying and estimating the number of mutual relationships or social circles in the population. Although these techniques provide a relatively inexpensive way of identifying individuals who are members of the intended study population, they may make it difficult to determine the extent to which the specific, sample-based study findings are representative of the larger population (Kalton, 1993; Thompson, 1997).
So-called adaptive sampling designs have been developed to sample populations that are "hidden" or difficult to reach and assess in a timely and resource-efficient manner or to estimate characteristics that are difficult to sample using conventional designs (e.g., the level of contamination in hazardous waste sites; the number and serostatus of men who have sex with men; the health status of commercial sex workers; condom use attitudes among undocumented migrant workers). With adaptive sampling, each step of the sampling process is determined by what the researcher has observed on the previous step (Thompson & Seber, 1996). Thus, in ascertaining the prevalence of a sexually transmitted infection in a particular area, public health officials might select and test individuals at random. (Some of the difficult ethical issues associated with STI research are considered in a later section of this chapter.) If one of these individuals tests positive, all of that person's friends and contacts are tested, and so on. As suggested by the research on social networks, adaptive cluster designs have particular appeal for studying drug use and sexually transmitted infections because the latter are not uniformly distributed within the population.
DATA COLLECTION
Data on the most accessible aspects of human sexuality-- attitudes, beliefs, intentions, norms, behaviors, practices, and functioning-- typically are obtained through face-to-face interviews, focus groups, self-administered questionnaires, computer-assisted personal interviews, and audio-enhanced computer interviews. With the exception of the recent computer-assisted methods, an extensive literature addresses the advantages and disadvantages of each of these tools (Clement, 1990; Coreil, 1995; Fetterman, 1998; Fowler, 1993; Gilbert, Fiske, & Lindzey, 1998; Jorgensen, 1989; Lavrakas, 1993; Locke & Gilbert, 1995; McCracken, 1988; Morgan, 1993; Ward, Bertrand, & Brown, 1991). Other sources of data on human sexuality include observational techniques, which have been used to study the sexual behavior of individuals, couples, and cultural groups; laboratory research, which has been used to test causal hypotheses; and clinical research, which has been used to study and assist individuals experiencing sexual problems.
Interview
Various interviewing methods have been used to study human sexuality (e.g., semistructured vs. structured; face-to-face vs. computer-assisted personal interview). The most famous example of an interview-based study may be the Kinsey reports on the sexual behavior of American males and females (Ellis, 1954; Kinsey, Pomeroy, & Martin, 1948; Kinsey, Pomeroy, Martin, & Gebhard, 1953). When data are obtained through face-to-face interview, the interviewer can respond quickly and appropriately upon hearing information that is inconsistent or incomplete. However, the use of interviews to obtain sexual behavior data, especially about sensitive behaviors such as extramarital relationships, has its disadvantages. In addition to being a time-intensive and relatively costly means of gathering data, a face-to-face interview may contribute to individuals reporting information that is biased or inaccurate. That is, individuals may respond in a socially desirable manner (Catania, Gibson, Chitwood, & Coates, 1990; Paulhus, 1991), and their responses may be affected by their perceptions of and reactions to the interviewer and/ or the interview situation (Benney, Riesman, & Star, 1956; Catania, Binson, Canchola, et al., 1996; Colombotos, Elinson, & Loewenstein, 1968; J. Freeman & Butler, 1976; Grimes & Hansen, 1984; T. Johnson, Hougland, & Moore, 1991; Kadushin, 1972; McBee & Justice, 1977; Shuman & Converse, 1968).
Imagine that you are a graduate student conducting a study of individuals' reactions to pornography. As part of the study, you show a pornographic movie to a group of mature undergraduate students. Suppose that the students like you and can identify with you because you look like one of them, and that they find the film enjoyable, arousing, and humorous. The next day, your professor repeats the study (i. e., shows the movie) with a different group of undergraduates. The professor is a no-nonsense social scientist and the students have long been intimidated by her. Unlike the students who viewed the movie on the previous day, these students find the movie distasteful and the whole experience anxiety producing. The different experiences of the two groups of students indicate that experimenter characteristics (such as age, gender, race/ ethnicity, clothing, and status) can affect individuals' reactions to sexually explicit movies (Abramson, Goldberg, Mosher, Abramson, & Gottesdiener, 1975; for related work on reactions to double entendres, see Abramson & Handschumacher, 1978).
A face-to-face interview constitutes a social interaction between two individuals and is therefore subject to the demands associated with most social situations. Unplanned social influences that are perceived and acted on by study participants are referred to as demand characteristics: participants encode cues that suggest something about the research hypothesis, infer the correct hypothesis or a hypothesis that results in the same behavior, and then act in a manner consistent with the inferred hypothesis (cf. Shrimp, Hyatt, & Synder, 1991). Demand characteristics are treated as an experimental artifact because they, and not the experimental manipulation, are responsible for the observed outcome. For example, in one study, women were either told or not told that the researchers were interested in studying the symptoms women experienced during the menstrual cycle (AuBuchon & Calhoun, 1985). After eight weeks of weekly assessment, the researchers found that "informed" women had reported more negative psychological and somatic symptoms at the premenstrual and menstrual phases of their cycle than did the remaining women. In this study, the two conditions associated with demand characteristics were met: the "informed" women thought they knew the purpose of the study, and they behaved accordingly. (Notice the similarity to the placebo effect that is often observed in medical trials.)
Interviewers and experimenters can also be influenced by their perceptions of both study participants and the study contexts. More important, experimenters' expectancies or beliefs about the likely reactions of study participants can bias the data they obtain (R. Rosenthal, 1978, 1980; R. Rosenthal & Rosnow, 1968; R. Rosenthal & Rubin, 1978). Although researchers cannot completely eliminate such influences, they should acknowledge them and try to reduce their impact by using a representative sample of experimenters (interviewers). (The rationale underlying the need for a representative group of study participants holds as well for a representative group of experimenters or interviewers.)
Focus Groups
A focus group consists of approximately 8 to 12 individuals who meet as often as necessary with a group facilitator (or moderator) to discuss a topic identified by the facilitator. (The purpose of the discussion is to generate information of principal use to the facilitator.) Focus groups provide researchers with an opportunity to study individuals' beliefs, attitudes, values, norms, and experiences within the context of a group interaction (Asbury, 1995; Carey & Smith, 1994; Knodel, 1995; Krueger, 1988; Morgan, 1988, 1993; Vaughn, Schumm, & Sinagub, 1996). In addition to the information generated during the group discussion, focus groups provide researchers with an opportunity to listen to and see how individuals talk to and interact with one another. As with other qualitative research methods, the data are transcribed, coded, and analyzed. Although many researchers have used focus groups in conjunction with in-depth interviews and questionnaires, researchers have also used focus groups to develop interviews and assist with the wording of questionnaire items.
The primary advantage of the focus group is its capacity to provide the researcher with insights into the psychological, social, economic, political, and cultural contexts that serve to define, promote, and constrain individuals' choices and behaviors. In particular, focus groups can provide researchers with several detailed and subtle perspectives in a relatively short period of time. As with any research tool, there are disadvantages associated with its use. First, the number of participants who are asked to speak for the study population is limited by necessity (and it is not too difficult to see how a selection bias could result in information that is not representative of the study population). Second, and more important, the group is subject to the same dynamics that can influence any other group. Discussion can be dominated by one or more individuals, especially if the moderator is not sufficiently experienced, and individuals may censor themselves or conform to the opinions expressed by other group members. Third, the use of multiple moderators at several sites may pose a strong challenge to the researcher's ability to achieve comparability within and across sites.
The use of focus groups as a research tool can be traced to work done in the 1930s. In recent years, the method has been used by researchers in different countries to study diverse topics in human sexuality, including the context of adolescent pregnancy in Nicaragua (Berglund, Liljestrand, Martin, Salgado, & Zelaya, 1997); women's knowledge and attitudes toward a proposed antenatal STI screening/treatment program in Haiti (Desormeaux et al., 1996); awareness and knowledge of sexually transmitted infections among married women living in rural Bangladesh (Khan, Rahman, Khanam, et al., 1997); child sexual abuse in Zimbabwe (Meursing, Vos, Coutinho, et al., 1995); sexual attitudes and views of male and female sexuality among married adults living in Bangkok (Knodel, Low, Saengtienchai, & Lucas, 1997); and sexual decision making among adolescent African American males (Gilmore, DeLamater, & Wagstaff, 1996).
Survey
To avoid the potential biases associated with face-to-face interviews, many researchers use self-administered questionnaires to collect sensitive information. By using a questionnaire, researchers seek to give respondents greater privacy when considering their responses to potentially embarrassing questions and greater flexibility in deciding when and where the information will be obtained. However, with this transfer of control to the respondent, the researcher may not be able to make appropriate adjustments when respondents report inconsistent data or decline to answer items. Moreover, with a self-administered questionnaire, some respondents may misunderstand the "skip patterns" designed to elicit information from specific individuals (e.g., "If 'No' go to Question 4a; if 'Yes' go to Question 4b"), and other respondents may not be able to read and comprehend some of the items.
To address concerns about the validity and reliability of information gathered through interview and self-administered questionnaire, a number of researchers have used telephone interviews (Bastani, Erickson, Marcus, et al., 1996; Catania, Coates, Stall, et al., 1992; Gibb, MacDonagh, Tookey, & Duong, 1997; Mishra & Serxner, 1994; Slutske et al., 1998) and computer-assisted data gathering (Bloom, 1998; Millstein & Irwin, 1983; Romer et al., 1997; Schneider, Taylor, Prater, & Wright, 1991). To varying degrees, these tools give respondents a greater sense of privacy and mitigate the social influence effects attributable to an interviewer's physical appearance or demeanor, although responses may still be influenced by the interviewer's vocal characteristics (Oksenberg, Coleman, & Cannell, 1986). Computer-assisted approaches are receiving increased attention because they permit individuals to respond privately while giving re-searchers the opportunity to (a) ask questions in a language spoken by the respondent; (b) ask the respondent additional questions if inconsistent or incomplete information is offered; (c) automate skip patterns and thereby ensure that the respondent is asked the appropriate questions; and (d) ensure that comparable information is gathered from respondents with various levels of reading skills.
Although the Kinsey Reports (Kinsey et al., 1948, 1953) were and still are recognized for their comprehensive look at the sexual behavior of U.S. males and females, most researchers appreciate the fact that the data were not representative of any larger population (W. Cochran, Mosteller, & Tukey, 1954; Turner, Danella, & Rogers, 1995). Indeed, until recently, much of what was known about sexual behavior was obtained from volunteers, clinical studies, and small samples. This situation has changed dramatically in the past thirty years, during which time a number of surveys of nationally representative populations (e.g., adolescents, young men and women, households) have been conducted. Typically, in these surveys a trained interviewer conducts face-to-face interviews with respondents in a private setting and a self-administered questionnaire is used to obtain the most sensitive data. Examples of surveys using this format include the 1971, 1976, and 1979 National Survey of Young Women (Zelnik & Kim, 1982); the 1979 National Survey of Young Men (Zelnik & Kanter, 1980); the 1988 and 1990 National Survey of Adolescent Males (Sonenstein, Pleck, & Ku, 1989); the 1991 National Survey of Men (Tanfer, 1993); and the 1992 National Health and Social Life Survey (Laumann, Gagnon, Michael, & Michaels, 1994). The National Health and Social Life Survey may be the most comprehensive survey of the sexual behavior of the general population that has been conducted to date (Laumann et al., 1994).
Examples of telephone survey include the population-based National AIDS Behavioral Surveys that were initiated in 1990 (Catania et al., 1992). The surveys obtained sexual behavior and HIV-risk data on a random sample of individuals residing in 23 U. S. metropolitan areas that accounted for a significant proportion of AIDS cases (e.g., Binson, Dolcini, Pollack, & Catania, 1993). In a recent example of the potential of computer-assisted surveys, researchers used laptops outfitted with an audio-enhanced, self-interviewing software program to obtain data from participants in the 1995 National Survey of Adolescent Males. Their findings suggested that estimates of sensitive and/ or illegal behaviors (e.g., male-male sexual contact, injection drug use, engaging in sex while drunk or high, engaging in sex with someone who injected drugs) obtained with a self-administered questionnaire may underestimate true prevalence of the behavior (Turner, Ku, Rogers, Lindberg, & Pleck, 1998).
Additional data on sexual behavior in the United States is provided through the General Social Survey, which has been conducted each year since 1972 (e.g., J. Davis & Smith, 1994; T. Smith, 1991); the recurring National Survey of Family Growth, which provides detailed information on the sexual and contraceptive behavior of women aged 15 to 44 (W. Mosher & Bachrach, 1995); the National Health Interview Survey, which has been fielded continuously since 1957 and provides information on the health of the civilian, noninstitutionalized U.S. population (National Center for Health Statistics, 1958, 1989); the National Longitudinal Survey of Youth, an annual study of individuals who were between the ages of 14 and 21 in 1979 (Center for Human Resource Research, 1995); and the biennial Youth Risk Behavior Survey of nineth-to twelfth-grade students conducted by the Centers for Disease Control and Prevention (Kolbe, 1990).
Direct Observation
There is, of course, a much more direct means than surveys or interviews for learning about people's sexual behaviors-- namely, observing them in the act. Observational studies, which sometimes include physiological measurements in addition to direct observation, are especially useful for obtaining information about the physical aspects of human sexuality. The most famous study to use this methodology is Masters and Johnson's (1966) investigation of the physiology of human sexual response. In their laboratory studies, they watched individuals and couples engaging in masturbation and intercourse, making detailed notes of observed physical changes (such as "sexual flush") and measuring physiological changes (such as penile circumference or vaginal lubrication). On the basis of these observations, they postulated that both men and women proceed through four distinct phases in their "sexual response cycles": excitement, plateau, orgasm, and resolution (Masters & Johnson, 1966). Although their theoretical integration of the physical data has been challenged (e.g., Robinson, 1976), Masters and Johnson are widely acknowledged for their pioneering use of observational and physiological measurement techniques.
Directly observed, laboratory investigations of human sexuality such as Masters and Johnson's are relatively rare. A more common observational technique is the ethnographic study, in which the researcher observes the population of interest (e.g., a society such as the Sambia or a social group such as injection drug users) in the natural environment. Ethnographic methods have historically been associated with the fields of anthropology and sociology, but are now employed by behavioral researchers of many stripes who wish to study sensitive human behaviors-- especially sex and illicit drug use-- in their naturally occurring contexts. This method is typically used to gather detailed descriptive data regarding the behaviors of the population of interest, rather than as a means of directly testing a specific hypothesis.
One of the keys to a successful ethnographic study is to minimize reactivity among the study participants. That is, their behavior should not be influenced in any way by the presence of the ethnographer. One approach requires that the ethnographer try to "blend into" the environment, so that the subjects of the study become accustomed to his or her presence and begin to react to it much as they would to any other feature of the natural environment. This is a passive, detached, and highly noninteractive ideal, and one that, like Star Trek's Prime Directive, is next to impossible to achieve. Instead, many ethnographers become participant-observers in the cultures they study. As participant-observer, the researcher becomes an acknowledged part of the culture, sometimes participating in the rituals or other activities of the group he or she is studying.
Usually, in ethnographies of sexual behavior, the emphasis is on observation rather than participation. Participation in general, and especially in studies of sexual behavior, raises questions of (a) reactivity among study subjects (i.e., is their behavior affected by the ethnographer's participation?); (b) the objectivity and validity of the ethnographer's account (can he or she objectively describe his or her own behavior, and is his or her experience typical of the culture or unique to his or her role as an outsider?); and (c) the ethics of engaging in or condoning potentially dangerous, illegal, or socially proscribed activities (such as illicit drug use, unsafe sex, or sex in public). These issues are highlighted in Humphries's (1970) famous ethnography of men who seek anonymous sex in public restrooms, Tearoom Trade. Conducting an observational study of men engaged in an illegal activity in a very small, confined space (public park restrooms, or "tearooms" in the argot of the men who frequent them) presented Humphries with a number of methodological difficulties to overcome. Initially, Humphries assumed the role of a "straight" (i.e., a man who enters the restroom solely to use the toilet) while attempting to observe the sexual activities in the tearooms, but as a straight, Humphries could remain in the tearoom for a short time only, during which his interest in the sexual goings became highly conspicuous. Worse, most sexual activity would cease as he approached the restroom. He soon realized that he needed to find a way to integrate himself into the tearoom trade subculture ("trade" refers to men who engage in homosexual activities but who do not self-identify as gay). For ethical reasons, Humphries rejected adopting an active sexual role and instead assumed the role of voyeur/lookout, or "watchqueen." In this role, Humphries was an accepted part of the tearoom sex scene (thus limiting the reactivity of the study population) and could freely observe the sexual behaviors of others, provided that he also performed his watchqueen duties by warning participants of the approach of straights, "chickens" (teenagers), and police officers (Humphries, 1970).
More generally, the anthropological literature makes clear that it is difficult to obtain accurate records of human sexual behavior (Abramson, 1992b; Abramson & Herdt, 1990; Herdt, 1981; Herdt & Stoller, 1989; Mead, 1961). Margaret Mead (1961), who is most famous for her investigations into adolescent sexuality in Samoa (Mead, 1928/ 1961), describes many of the difficulties. She notes as a starting point that human sexual behavior is intimately linked with privacy. As a result, there is an inherent tension between the individual's desire to be protected from unsolicited intrusion and the researcher's desire to make a public record that can be examined by others (Abramson, 1990; Abramson & Pinkerton, 1995; Mead, 1961). Similarly, Mead notes that sexuality is characterized in almost every society by gaps in awareness (e.g., accurate descriptions of genitalia) and specific taboos (e.g., homosexuality) that contribute to an inability or unwillingness to talk about one's sexual behavior. Finally, Mead asserts that there is no cultural rationale for describing honestly one's sexual behavior and that there are discrepancies between institutionalized statements of what is appropriate sexual behavior and actual sexual behavior and practices.
In the past thirty years, more methodological concerns have been raised regarding cultural prohibitions and unconscious processes that make the assessment of sexual behavior an extraordinarily difficult objective. For example, Abramson and Herdt (1990) describe two levels of bias in the cross-cultural assessment of sexual practices. The first level is related to the language used to identify culturally sensitive categories. Is there a generic category for "prostitute," or is the category gendered? (Compare "hustler" for male to "hooker" for female.) How may these labels be used, in what contexts, and by whom? More important, to what extent do situation-specific meanings and behaviors bias individuals' responses to questions about sexual behavior?
The second level of bias is related to the manner in which sexuality-related ethnographic information is collected, and in particular, with interviewing protocols. When one considers the linguistic context and the need for culturally sensitive procedures, the question of where individuals should be interviewed arises immediately. Should the interview be conducted in public or in private? Should the individuals conducting the interviews be of the same culture, gender, social class, and sexual orientation as the interviewee? An ethnographer usually works alone and establishes a close, personal relationship with the social group members participating in the study. The ethnographer will learn their language and customs, participate in the group's activities, and perhaps live with them. Although these practices undoubtedly enhance trust and facilitate understanding, the literature on interviewer/experimenter effects and the fierce debates about the impact of specific anthropologists (e.g., Margaret Mead; see D. Freeman, 1983) argue forcefully for the use, when possible, of multiple investigators with different sociodemographic characteristics.
STUDY OF SOCIAL NETWORKS
One area that has seen an explosive growth in the use of ethnographic methods is the study of social networks. A social network analysis focuses on the relationships among social entities or actors and the patterns and implications of those relationships (Wasserman & Faust, 1994). More important, it focuses on the interactions among the actors, as opposed to the attitudes, beliefs, intentions, and behaviors of an individual actor. This perspective has grown in popularity because many important aspects of social life are conducted by individuals organized into social networks (Galaskiewicz & Wasserman, 1994). It has become increasingly important to ethnographers studying urban life because an understanding of the social network is necessary to obtain information on less public events and establish relationships with key informants ( J. Johnson, 1994). The social network perspective has also grown in importance because it is well suited to studying the injection drug use and sexual networks that transmit HIV and other sexually transmitted infections (Anderson & May, 1991; Morris, 1994; Neaigus, Friedman, Curtis, et al., 1994; Rhodes, Stimson, & Quirk, 1996; Rothenberg, Potterat, & Woodhouse, 1996). In particular, a social network perspective recognizes the fact that, like many material goods, drugs and infections are exchangeable and, as such, travel routes that connect interdependent social networks (e.g., from individual transactions in shooting galleries to mass distribution by drug cartels). In addition, and unlike many epidemiological models, a social network perspective recognizes the fact that individuals do not select their drug and/or sexual partners at random: sociodemographic characteristics strongly influence interaction opportunities and define the appropriate groups from which one can select one's friends, neighbors, associates, and partners.
Ethnographers have made a number of important contributions to the study of human sexuality and sexual behavior. Their work continues to show that public proclamations (surface-level material) can often be unreliable; that subtle methods can yield detailed information about the social experiences of individuals and their social networks; and that rapport and intimate understanding are critical (Abramson, 1992a).
CLINICAL RESEARCH
Clinical research, naturally enough, is research that is conducted in a clinical setting (e.g., a hospital, clinic, or physician's or therapist's office); with a clinical population (e.g., men with erectile dysfunction or postoperative transsexuals); or that concerns clinical issues related to sexuality (e.g., physical aspects of sexual satisfaction). The most frequently studied clinical issues concern the diagnosis and treatment of difficulties with specific aspects of sexual attraction, arousal, and performance.
Clinical research often adopts a case study approach, in which one or more persons with the condition of interest are described in detail. For example, Coleman and Bockting (1988) describe a 36-year-old female-to-male transsexual who was "heterosexual" prior to sex reassignment surgery and "homosexual" after it-- that is, who was attracted to men both before and after surgery. This case study illustrates the separability of the concepts of gender identity and sexual orientation, in that the patient sought surgery to bring her (female) anatomical sex into alignment with her (male) psychological gender, even though the surgery also transformed her from a "straight" woman into a "gay" man. Importantly, it also demonstrates a critical scientific function of case studies: if properly documented, they can serve as disconfirmatory counter-examples to prevailing theories (e.g., that gender identity and sexual orientation are inexorably connected). Moreover, case studies can offer deep insight into the causes and correlates of the study condition and thereby can help re-searchers generate testable hypotheses (Abramson, 1992a). Although the generalizability of case study findings is limited, the case study approach makes up in specificity what it lacks in generality.
Large-scale clinical studies are often undertaken to evaluate the efficacy of new treatments for various forms of sexual dysfunction, or to assess the relationship between sex and other health-related issues. As an example of the latter category, a recent study published in the prestigious British Medical Journal suggests that maintaining an active sex life can have beneficial effects on life expectancy; specifically, the authors found evidence of a dose-response relationship, such that men with greater orgasmic frequency exhibited a significantly lower risk of death from coronary heart disease (Davey Smith, Frankel, & Yarnell, 1997).
Erectile dysfunction is also commonly associated with other medical conditions, such as hypertension, diabetes, and depression, as well as with psychogenic factors. In 1998, the Food and Drug Administration (FDA) approved a new drug for the treatment of erectile dysfunction on the basis of 21 randomized, double-blind, placebo-controlled trials (Lamberg, 1998). According to the FDA, the new drug, Viagra (oral sildenafil), led to at least some improvement in 7 of 10 men with erectile dysfunction, compared to 2 of 10 men who reported improved functioning when given a placebo. Surprisingly, perhaps, the clinical efficacy trials relied on a self-report measure, the International Index of Erectile Dysfunction, rather than on direct physiological measurement of erectile function, such as nocturnal erections (Lamberg, 1998).
More generally, the fourth edition of the venerable Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) (American Psychiatric Association [APA], 1994) defines sexual dysfunction as a problem characterized by a disturbance in both sexual desire and the psychophysiological changes that precipitate sexual arousal or orgasm. Such problems are often accompanied by distress and relationship difficulties. They are also surprisingly common: in one study of "happily married couples," 40% of the men reported erectile or ejaculatory dysfunction and 63% of the women reported arousal or orgasmic dysfunction (Frank, Anderson, & Rubinstein, 1978). From a diagnostic perspective, research tends to focus on precipitating factors (e.g., relationship conflict, medical cofactors) and delineating specific syndromes (e.g., sexual desire disorders) (H. Kaplan, 1979; Stuart, Hammond, & Pett, 1988; Zilbergeld & Kilmann, 1984). Therapy research tends to focus more on developing specific treatment formats (e.g., masturbation training for female orgasmic dysfunction) that are drawn from different treatment modalities (e.g., behavioral, cognitive, marital, psychodynamic) (Masters & Johnson, 1970; Rosen & Leiblum, 1992; Schover & LoPiccolo, 1982; Zilbergeld, 1992).
Another main area of clinical research concerns the diagnosis and treatment of sexual attraction disorders. For many years, homosexuality was conceptualized as a psychiatric problem because it represents a deviation from the statistical norm in terms of sexual attraction (i.e., unlike most of the population, homosexuals are sexually attracted to members of their own sex). Homosexuality was thus considered deviant. The larger question, however, is whether this "deviation" encompasses identifiably pathological psychological processes, or whether sexual attraction is simply a form of diversity, like skin or eye color, or a preference for brunettes over blondes.
Although political and religious issues did (and continue to) play a significant role in this debate, the scientific question of whether homosexual attraction is pathological was largely settled in the 1950s and 1960s by researchers such as Evelyn Hooker. Hooker (1957) administered a series of psychological tests to nonclinical samples of homosexual and heterosexual populations and demonstrated that there were no significant differences between the two populations. Subsequent research supported this finding, which precipitated a major change in how homosexuality was viewed. In the 1970s, the American Psychiatric Association passed a resolution that stated, "homosexuality per se implies no impairment in judgment, stability, reliability or general social or vocational capabilities" and removed homosexuality from its list of mental disorders (Allgeier & Allgeier, 1995).
At present, most sexual attraction disorders are characterized as paraphilias, which are defined as intense and recurring sexual feelings or behaviors involving nonhuman objects, suffering or humiliation, or nonconsensual activities (e.g., exhibitionism or acts involving children). The treatment of these disorders is also a focus of clinical research. Behavioral and cognitive therapies are the preferred technique (e.g., Marshall, Eccles, & Barbaree, 1991), although there is considerable debate about the effectiveness of treating sexual offenders, given the high recidivism rates that are often observed. Hormonal and other pharmacological therapies have shown some promise in this regard (Bradford, 1998; Gijs & Gooren, 1996; Rösler & Witztum, 1998). Related areas of treatment research also include improving techniques for counseling survivors of sexual assault and designing therapies for the nonoffending partner or parent in cases of incest (Finkelhor & Berliner, 1995). There is also an ongoing debate over the status and possible treatment of "sexual compulsivity" or "sexual addiction" (Abramson & Pinkerton, 1995; Coleman, 1991; Levine & Troiden, 1988; Quadland, 1985).
LABORATORY RESEARCH
Researchers conduct laboratory studies of human sexuality to test hypotheses about causal relationships. For example, Kelley, Byrne, Greendlinger, and Murnen (1997) used the context, sex of the viewer, and seven dispositional variables to predict heterosexual college students' affective and evaluative responses to three types of sexually explicit film; Murnen, Perot, and Byrne (1989) studied situational determinants of and individual differences associated with female college students' coping reactions to unwanted sexual advances; and Przybyla and Byrne (1984) examined male and female college students' recall of and their affective and sexual/physiological reactions to erotic stimuli in the presence of auditory and visual distractions.
Researchers use laboratory settings because such settings give them greater control over conditions that could account for the hypothesized relationship. In particular, laboratory settings give researchers the ability to control the number, magnitude, and type of influences, as well as to constrain the manner in which these influences exert their effects on the causal relationship of primary interest. In turn, this control enhances the researcher's ability to conclude, when supported by the data, that the assumed cause and effect covary, and that this covariation cannot be attributed to alternative causes. Because the assumed cause precedes the effect in time, covaries with the effect, and is the most plausible explanation of that covariation, the researcher concludes that the hypothesized causal relationship holds. This is the logic behind studies conducted in laboratory settings.
On occasion, the nature of the research is such that the laboratory is the only place where it can be conducted. Studies of sexual response to stimuli spring to mind immediately. For example, Janssen, Vissenberg, Visser, and Everaerd (1997) conducted a study to compare two types of penile strain gauge (devices used to measure male sexual arousal). Laan, Everaerd, and Evers (1995) studied female sexual arousal to visual stimuli that were sexual, anxiety-inducing, sexually threatening, or neutral. The researchers measured vaginal pulse amplitude and blood volume, skin conductance, and heart rate-- measurements best taken in the laboratory setting. However, even when instruments are not required to obtain data, researchers may want to take advantage of the privacy, quiet, and safety the laboratory provides.
When conducting a laboratory study, a researcher's primary concern has often been with the study's internal validity-- the judgment that the observed statistical relationship was a true causal relationship (Cook & Campbell, 1976). Internal validity is a "diagnosis by exclusion." Confidence in the judgment that the study has high internal validity is achieved by eliminating (preferably by design and not by subsequent argument) the plausible challenges to the putative causal relationship. However, with an increased involvement in the body politic (e.g., social action research, applied social research, evaluation research, policy research) and with growing pressure to provide relevant answers to social problems (race relations, economic disadvantage, adolescent pregnancy, welfare dependency, violence and crime, HIV infection), researchers have given more attention to designing and analyzing studies that have more external validity. A finding from a study with high external validity is one that holds across populations and settings (Cook & Campbell, 1976).
Although laboratory studies have enhanced our understanding of human sexuality, questions have been raised about the generalizability of the findings obtained thereby. These questions reflect concerns about the nature of the set-ting and the task experienced by the study participants (i.e., concerns about the study's ecological validity). The questions also reflect concerns about the use of volunteers. Evidence suggests that the individuals who choose to participate in studies of human sexuality differ from those who do not choose to do so (Barker & Perlman, 1975; Bogaert, 1996; Kaats & Davis, 1971; Morokoff, 1986; Strassberg & Lowe, 1995; Wolchik, Braver, & Jensen, 1985; Wolchik, Spencer, & Lisi, 1983). Strassberg and Lowe found that undergraduate sex research volunteers reported more positive attitudes toward sexuality, less sexual guilt, and more sexual experiences than nonvolunteers. Bogaert found that undergraduate male volunteers were more sexually experienced, more interested in sexual variety, and more erotophilic. In addition, volunteers were higher in sensation seeking and lower in social conformity than nonvolunteers. Morokoff found that unmarried female undergraduate volunteers had more noncoital sexual experience, more masturbatory experience, and less sexual inhibition than nonvolunteers. These and similar findings pose a serious threat to the external validity of study findings that are based on volunteers. Future research on human sexuality should make an effort to include more diverse populations and should take greater care to control for possible selection bias.
FIELD STUDIES
Field studies are defined by their location: they are research studies conducted in the field (as opposed to a research laboratory). Examples of field studies include the HIV prevention intervention developed for mentally ill homeless men by Ezra Susser and his colleagues (Susser et al., 1998; Susser, Valencia, Miller, et al., 1995; Susser, Valencia, Sohler, et al., 1996; Susser, Valencia, & Torres, 1994). The intervention consisted of 15 sessions, which is more sessions than many other prevention interventions. Contributing to the risk of infection were cocaine and injection drug use, and unprotected sex. The intervention combined social cognitive theory skills training (i.e., role play, modeling, social reinforcement, and feedback) with various clinical approaches and activities. The latter were centered on the kinds of conversations, games, and activities that were routine to shelter life. Other researchers have offered similar cognitive-behavioral interventions to prevent sexually transmitted HIV infection among adolescents ( Jemmott, Jemmott, & Fong, 1992, 1998; St. Lawrence, Brasfield, Jefferson, Alleyne, & O'Bannon, 1995); gay men (Kegeles, Hays, & Coates, 1996; Kelly, St. Lawrence, Betts, Brasfield, & Hood, 1990; Kelly, St. Lawrence, Diaz, et al., 1991; Kelly, St. Lawrence, Stevenson, et al., 1992); and economically disadvantaged women (Kelly, Murphy, Washington, et al., 1994).
Field studies, prevention interventions, and planned program evaluations share many of the strengths and weaknesses of laboratory-based studies. Indeed, many graduate students in education and the social sciences receive training in the logic of experimental design, if they are not actually trained to conduct lab-based experiments. Thus, it should come as no surprise that field studies and prevention interventions rely on convenience samples. However, the evidence for volunteer bias in lab-based studies (e.g., Bogaert, 1992, 1996; Griffith & Walker, 1976; Morokoff, 1986; Nirenberg, Wincze, Bansal, et al., 1991; Strassberg & Lowe, 1995; Wolchik et al., 1983, 1985) suggests that the individuals who choose to participate in a prevention program differ from those who do not choose to do so. Prevention interventions may have high internal validity; however, the failure to address concerns about their external validity will result in their having less value to programmers and policymakers, and less support from the larger community.
MEASURES
It is often said that the fundamental activity that distinguishes science from other human pursuits is the taking of measurements. Indeed, it has been argued that without the successive refinement in measurement, scientific progress would cease and science would grind to a halt (Kemeny, 1959). With this emphasis on measurement, scientists have an obligation to demonstrate that their measures are valid and reliable. Taken once, the measure should provide an accurate value; taken repeatedly, it should provide consistent values.
SEXUAL BEHAVIOR AND SELF-REPORTS
Researchers who study human sexuality need to be particularly concerned about the validity and reliability of their measures because they often rely on individuals' reports of their sexual experiences. These self-reports reflect the experiences that individuals can recall and choose to share. Thus, when re-searchers study aspects of human sexuality that are socially mediated, they need to take into account the capacities and limits of human memory and the conscious and unconscious motivations that shape self-reports. In all fairness, all researchers who rely on self-report should proceed cautiously (even when the studied phenomenon is as public as weight gain; see Bowman & DeLucia, 1992).
Numerous studies and reviews have considered the validity and reliability of self-reports of sexual behavior. Many of the early studies reflect research on the sexual behavior of adolescents and young adults. The finding that adolescent, young adult, and adult males in the United States, Britain, France, New Zealand, and Norway report many more sex partners than their female counterparts has been one of the most consistent and troubling research findings (Wiederman, 1997). The suggested explanations for the discrepancy include some form of sampling bias (e.g., males have greater sexual contact with individuals such as young females or prostitutes who are not included in the study sample), response bias (e.g., differences in the way males and females define "sex" and "sex partner"), and/or outlier influence (i. e., men who report large numbers of partners may inflate the mean) (Morris, 1993; Wiederman, 1997).
More recent studies of the validity and reliability of self-report data reflect research on risk behaviors for sexually transmitted infections, specifically HIV infection (Abramson, 1988; Abramson & Herdt, 1990; Alexander, Somerfield, Ensminger, et al., 1993; Catania, Gibson, Chitwood, et al., 1990; Catania, Gibson, Marin, Coates, & Greenblatt, 1990; Coates, Calzavara, Soskolne, et al., 1988; Coates, Soskolne, Calzavara, et al., 1986; D. Cohen & Dent, 1992; Jaccard & Wan, 1995; James, Bignell, & Gilles, 1991; Kauth, St. Lawrence, & Kelly, 1991; McLaws, Oldenburg, Ross, & Cooper, 1990; Padian, Aral, Vranizan, & Bolan, 1995; Saltzman, Stoddard, McCusker, Moon, & Mayer, 1987; Turner & Miller, 1997; Upchurch, Weisman, Shepherd, Brookmeyer, et al., 1991; Zenilman, Weisman, Rompalo, et al., 1995). Concerns about the reliability of the data collected by HIV researchers are heightened because the behaviors that result in HIV transmission are illegal (injection drug use), stigmatized (anal intercourse), or difficult to enact (getting a reluctant partner to use condoms). On the whole, researchers have concluded that self-reports of sexual behaviors are reliable, particularly when the recall period is short and individuals are asked to recall salient behaviors. However, an uncritical treatment of individuals' reports of sexual behavior may be unwarranted.
In a recent experiment (Berk, Abramson, & Okami, 1995), half of the participants were asked to keep a daily diary for two weeks; the remaining participants were not. Within each group, participants were assigned to a memory enhancement, placebo enhancement, or no instruction condition. The purpose of the study was to provide an estimate of the error in the recall of (recent) sexual behavior, examine factors that might facilitate recall on a survey of sexual behavior, and provide further insight into how people experience and encode human sexual behavior. Although the study was weighted toward remembering recent sexual experiences (in that the recall period was the past two weeks and participants were explicitly encouraged to remember events as accurately as possible), the most striking finding was how little participants could remember about their sexual experiences. This was true even for the participants who kept a daily diary and wrote about their experiences. Furthermore, the nondiary participants reported more sexual behavior than did the diary participants, suggesting that individuals tended to inflate their sexual experiences without the help of a diary.
Why? Do the findings reflect error or chance, or do they suggest something about the processes associated with the encoding and recall of sexual experiences? The authors of the study believe the latter and have called for further study of the recall of sexual experiences as an adjunct to the assessment of sexual practices (Berk et al., 1995).
PHYSIOLOGICAL MEASURES
As we have discussed throughout this chapter, there are many obstacles to collecting valid and reliable data on human sexual behavior, including reliance on retrospective recall, social desirability effects, and experimenter effects. Conversely, there are several characteristics that facilitate the measurement of human sexuality. Unlike many psychological variables (e.g., guilt, depression), sex is often an overtly expressed behavior, which can be counted (e.g., "How often do you masturbate per month?") or measured physiologically (e.g., erectile tumescence). The latter is particularly significant because physiological assessment can provide several advantages in theory, measurement, and objectivity.
One of the most interesting physiological measurement techniques makes use of heat to indicate changes in sexual status. All objects-- including humans-- emit infrared energy as a function of temperature. This infrared energy (or "heat") can be measured through a process known as thermography. Thermography is a noninvasive means of detecting and photographing individual heat-generation patterns to indicate physiological condition and functional changes within (M. Bacon, 1976). Thermography is also a very useful physiological measure for studying sexual arousal in humans. Because pelvic vasocongestion and myotonia (increased blood flow and muscular spasms, respectively), both of which are associated with increased temperature, are the principal peripheral physiological responses that accompany sexual arousal, thermography is an excellent way of documenting the blood flow rate and transfer of heat that underlie the experience of being sexually aroused (Abram-son, Perry, Rothblatt, Seeley, & Seeley, 1981; Abramson, Perry, Seeley, Seeley, & Rothblatt, 1981; Seeley, Abramson, Perry, Rothblatt, & Seeley, 1980). Thermography has also been useful in documenting an asymmetrical vasocongestive pattern in the pectoral region that accompanies the sexual response cycle (Abramson & Pearsall, 1983). Thus, unlike many paper-and-pencil social science measures, thermography represents a good match among object (sexual arousal), theory (heat distribution), and measurement scale (temperature).
A number of other physiological measures are useful for studying human sexual arousal. Probably the best-known examples are the penile strain gauge and the vaginal photoplethysmograph. The former can measure minute changes in penis size (including each pulse of blood in the penis), and the latter measures increased vaginal blood volume. These measures have been used, for example, to study men's and women's physiological reactions to sexually explicit films or pictures (Rosen & Beck, 1988). Of course, researchers need to be particularly sensitive when their experiments involve direct genital measurement (Abramson, Perry, Rothblatt, et. al., 1981). Using same-gendered experimenters, providing privacy to participants, and allowing self-placement and monitoring of genital devices should be employed when possible.
Besides genital blood flow, new technologies, such as magnetic resonance imaging (MRI), offer the possibility of examining other bodily responses that are instrumental to sexual arousal, including the role of the brain. Additionally, endocrinological studies of the hormonal changes that accompany or influence sexual behavior are very important and active areas of research (Meyer-Bahlburg, 1995; J. Wilson, 1995).
PERSONALITY MEASURES
There are several measures of specific personality traits that are germane to the topic of sex research; these be discussed momentarily. First, however, we discuss the relationship of sexual behavior to the comprehensive personality model of Hans Eysenck. (In this exposition, personality is viewed as an essentially stable component of the psyche that determines fundamental patterns of interaction between self and environment.)
Like many other contemporary personality theorists, Eysenck (1947; Eysenck & Eysenck, 1969) posited that personality could be factored into a small number of nearly orthogonal dimensions, or traits. The three dimensions proposed by Eysenck were (1) extroversion, a nonspecific mixture of sociability and impulsivity; extroverts are sociable, lively, active, assertive, and highly sensation seeking; (2) neuroticism, largely a measure of emotional instability; persons who are high in this trait tend to be anxious, depressed, tense, and to suffer from feelings of guilt and low self-esteem; and (3) psychoticism, which reflects asocial and atypical attitudes and tendencies, including cruelty, indifference to the feelings of others, and paranoia; highly psychotic individuals are characteristically aggressive, cold, egocentric, impersonal, and impulsive. Relative standing on these three dimensions (E-P-N) is assessed using the Eysenck Personality Questionnaire, together with a "lie scale" that is used to detect dissimulation and social conformity (Eysenck & Eysenck, 1975). (The labeling of Eysenck's three factors as extroversion, neuroticism, and psychoticism is misleading. For example, a high score on the psychoticism scale does not indicate psychosis in the clinical sense; nor does a high neuroticism score necessarily reflect neurosis in the Woody Allen sense.)
In his book, Sex and Personality, Eysenck (1976) made a number of concrete predictions regarding the relationships of E-P-N factors to sexual behavior, especially focusing on the expected correlation between extroversion and enhanced sexuality (Barnes, Malamuth, & Check, 1984). Empirical tests of these predictions suggest the following (see Abramson, 1973; Barnes et al., 1984; Eysenck, 1976; Giese & Schmidt, 1968). First, compared to introverted people, extroverted individuals are more likely to be sexually promiscuous and hedonistic; desire greater sexual variety; are more likely to have multiple partners; and experience greater enjoyment and satisfaction from engaging in conventional sexual activities. Those who are higher on the psychoticism scale are more likely to exhibit high libidos and to hold favorable attitudes toward unconventional sexual activities, including impersonal sex, premarital and extramarital sexuality, and coercion. High neuroticism scores are generally associated with increased guilt, decreased sexual enjoyment, and intolerance of premarital sex. The relationships of these and other personality variables (such as sensation seeking) to HIV risk taking is examined in Pinkerton and Abramson (1995).
One of the main drawbacks to the use of a comprehensive personality inventory such as Eysenck's is that it is too general and too comprehensive. What is needed in most research situations is a specific measure of the relevant sexuality construct. For example, a study of masturbation guilt requires an instrument that is devoted to assessing masturbation guilt. Fortunately, a number of specific measures exist, including ones for masturbation attitudes and guilt (Abramson & Mosher, 1975); sexual guilt and repression (D. Mosher, 1966, 1973); and erotophobia-erotophilia (Byrne, 1983; Byrne & Schulte, 1990), which measures attitudes toward sex-related materials and activities. A wealth of research utilizing these measures has established their utility (see, e.g., Abramson & Handschumacher, 1978; W. Fisher, 1980; Galbraith & Mosher, 1968; D. Mosher & Abramson, 1977; Schill & Chapin, 1972; Schwartz, 1973).
The best-known instruments for assessing sexual orientation and gender identity are the Kinsey scale and the Bem Sex Role Inventory, respectively. Kinsey measured sexual orientation along a 7-point continuum ranging from strictly heterosexual (0) to strictly homosexual (6), with various mixes of opposite-sex and same-sex sexual behaviors falling between these extremes (e.g., the midpoint of Kinsey's scale describes someone who engages in equal amounts of heterosexual and homosexual activity). However, sexual behavior is not necessarily congruent with sexual attraction or with sexual identity. For this reason, contemporary research often uses multiscale instruments that separately measure these different aspects of sexual orientation.
Rather than juxtaposing masculinity and femininity along a bipolar continuum, as earlier instruments had, the Bem Sex Role Inventory (BMSI) assesses them separately (Bem, 1974). The original form of the BMSI asks the respondent to rate himself or herself on 20 stereotypically masculine personality traits (e.g., assertive, dominant), 20 stereotypically feminine traits (e.g., com-passionate, affectionate), and 20 filler items. (Later, a short form was devised that consists of half the number of items on the original.) Because the BMSI measures femininity and masculinity separately, a person can appear both highly masculine and highly feminine; such a person is considered androgynous (at the other extreme, someone who is neither masculine or feminine is labeled undifferentiated) (see, e.g., Ballard-Reisch & Elton, 1992; Hyde, Krajnik, & Skuldt-Niederberger, 1991; Taylor & Hall, 1982).
PSYCHOSOCIAL MEASURES
Although some researchers develop psychosocial scales and instruments for each new study they conduct, a number of these tools have been collected in book form and are available off the shelf for use by other investigators. For example, C. Davis, Yarber, Bauserman, Scheer, and Davis (1998) discuss more than 200 measures of sex-related states, traits, behaviors, and outcomes that have been used in research and clinical settings. In most instances, the instrument is provided along with information on its purpose, timing, scoring, interpretation, and psychometric properties (reliability, validity). The measures address over 50 areas, representing research on abortion, abuse, aging, arousal, attitudes, coitus, contraception, dysfunction, esteem, fantasy, gender identity, gender roles, harassment, homophobia, homosexuality, intimacy, molestation, orgasm, rape, relationships, sexual history, sexual risk, sexually transmitted infections, stereo-types, transsexualism, and vasectomy.
Numerous general social psychological measures can be found in Robinson, Shaver, and Wrightsman (1991). With one notable exception, the topics and measures are not specific to research on human sexuality. However, several of the theories underlying current behavior change and prevention interventions require that researchers use measures that are similar to the measures found in this reference.
ETHICAL ISSUES
Two issues arise immediately when considering the ethics of research on human sexuality. The first issue reflects the varied opinions regarding what is ethical, the nature of the information on human sexuality that individuals need, and the manner in which this information is to be obtained and shared with them. There are many opinions on these matters and much of what is offered is important. However, few research studies have addressed the under-lying questions explicitly. That is the second issue. Although many sex re-searchers have to address ethical problems when they plan and conduct their studies, they seldom consider ethical problems per se to be an important focus for research.
Consider the question of who should participate in human sexuality studies. Presumably, we want participants of all races/ethnicities, both genders, all sexual orientations, and different age groups. However, should all participants be 18 or older, or can they be younger? Many of the critical questions about human sexuality have to do with its development. For example, some researchers are interested in determining when individuals become aware that they have sexual feelings for others. Although we have retrospective data and parental observations on this process, we rarely get firsthand information from children. Should we settle for retrospective data, or should we develop ethical methods that allow us to obtain data from children? Even when there is agreement on the question of who should participate, we may have concerns about the appropriateness of assigning treatments to individuals at random. The issues are particularly difficult when individuals in need are to receive a placebo, a "treatment" that is known to have no effect. A placebo is often used in a randomized trial when the researcher wants to determine if an active treatment impacts some outcome. For present purposes, the active treatment could be a new drug, a pregnancy prevention program, or an HIV prevention intervention. In the absence of any preexisting differences between treatment and control groups, data from a control group provide information on the outcomes that would have been observed among treated individuals in the absence of any treatment effect. Randomization bolsters the study's internal validity: it ensures that any differences observed prior to treatment are due to chance and not due, for example, to a selection bias exercised by the study participant, referring physician or therapist, social worker, or researcher. The use of placebo receives less challenge when there is no broadly accepted standard treatment that can serve as a control and the consequences of receiving an inert treatment are mild. Its use is challenged more when prior studies indicate that something may work for some individuals under some conditions and it is clear that the consequences of receiving an inert treatment are serious.
Different members of society also hold different opinions on the kinds of sex-related information and materials that should be available and how they should be disseminated. School-based sex education and condom distribution programs illustrate some of the current challenges. Although these programs are designed to reduce the risk of adolescent pregnancy and sexually transmitted infections, some critics argue that they violate religious principles or usurp a parental role and responsibility; others argue that they encourage children to be sexually active (the available evidence suggests that they do not; see Grunseit, Kippax, Aggleton, Baldo, & Slitkin, 1997; Kirby, 1984, 1985; Kirby, Short, Collins, et al., 1994; Kirby, Waszak, & Ziegler, 1991; Stout & Kirby, 1993). Because data can only address some of the concerns (e.g., whether education and condom distribution programs encourage youth to be sexually active, or whether parent-child relationships are damaged if trained teachers provide instruction on human sexuality), researchers, practitioners, and teachers must learn to communicate more effectively with parents and other community members.
Communicating with individuals about the purpose and nature of the study, as well as the risks and benefits of participation, is a necessary requirement for obtaining individuals' informed consent. Additionally, when researchers provide individuals with the information needed to make an informed decision, the latter are less likely to question researchers' motives and/ or behavior. Clearly, individuals and the communities asked to host and support sex research are more likely to question a researcher's motives and behavior when they learn that they have been deceived. The infamous Tuskegee syphilis study, in which medical treatment was intentionally withheld from some southern African American men in order to study the course of untreated syphilis, may be the most well-known example of scientific misconduct involving multiple levels of deception, long-term efforts to see that participants did not receive treatment, and a legacy of mistrust and suspicion (Brandt, 1978; Silver, 1988; Thomas & Quinn, 1991). Although institutional review boards now operate to safeguard the rights of research participants and ensure that they receive adequate protection, researchers retain primary responsibility. Researchers should be as clear as possible about the nature and purpose of the study. Most important, they should make sure that individuals understand all of the information contained in the informed consent form before signing, and they should provide participants with a detailed debriefing (Perry & Abramson, 1980). When provided with such, the evidence indicates that individuals often find participating in sex research to be an interesting and enriching experience (Abramson, 1977).
Proper graduate training should be designed to equip sex researchers with sensitivity to others as well as the knowledge needed to conduct ethical research. However, another way to ensure that participants are treated ethically in sex research is to enhance their sexual literacy. The better informed participants are about sex, the more likely they are to participate in a knowledgeable and conscientious manner. In Cultural Literacy: What Every American Needs to Know, Hirsch (1987) argued that Americans were no longer culturally literate, and ignited an enduring controversy about the province and effectiveness of American education, formal and otherwise. We believe the same is true for "sexual literacy" (Abramson & Pinkerton, 1999). Being informed about sexual matters serves to ensure that sexual liabilities do not exceed potential benefits. As we have argued elsewhere (Abramson & Pinkerton, 1999), improving the sexual literacy of Americans is an important and laudable goal. To do this would require, at a minimum, that society begin to better support the structures that create sexual literacy, that is, sexual research and scholarship and the teaching of sexuality in schools, from elementary school through the university.
CLINICAL APPLICATIONS
Research on human sexual behavior has numerous clinical applications. The most obvious is the study of sexual dysfunctions (e.g., the inability to become aroused, achieve orgasm, or maintain an erection). Assessment of the type of dysfunction is one clinical application; treatment is another. Sex therapy is the primary method used to treat many sexual dysfunctions. However, the extraordinary attention recently given to the anti-impotence drug Viagra suggests that the present focus will shift increasingly to various forms of drug therapy. Indeed, a number of companies are investing a considerable amount of re-search effort and enormous amounts of money to develop drugs that enhance sexual performance and eradicate sexual dysfunction in men and women.
The study of sexual satisfaction in relationships is another clinical application of research on human sexuality. Marriage is an obvious focus for these studies (Perlman & Abramson, 1982). However, sexual satisfaction is equally important to elderly individuals (Kellett, 1991; Leiblum, 1990; Schiavi, 1990; Schlesinger, 1996; Segraves & Segraves, 1995); gay men, bisexuals, and lesbians (Dennen, Gijs, & van Naerssen, 1994; Rose, 1994; Rosenzweig & Lebow, 1992; Rosser, Metz, Bockting, & Buroker, 1997); disabled individuals (Cole & Cole, 1993; Mona, Gardos, & Brown, 1994; Pitzele, 1995; Sipski & Alexander, 1997); and individuals who are or are not married (Byers, Demmons, & Lawrance, 1998; K.-A. Lawrance & Byers, 1995; MacNeil & Byers, 1997). In fact, as a basic clinical issue, sexual satisfaction has enormous significance in terms of self-esteem and psychological health for most people, from adolescence onward.
Clinical applications also come in other forms. Because we tend to be sexually attracted to people whom we find physically attractive, studies of physical attractiveness can yield findings that have important implications for clinical practice (Byrne, 1997; E. Smith, Byrne, & Fielding, 1995). To examine the question of physical attractiveness in the context of psychotherapy, Murray and Abramson (1983) created a fictitious college student medical history form using gender-neutral writing and a set of psychological problems (e.g., lack of assertiveness, uncertainty about career choices). Four forms were created: two were purportedly written by males, and two were purportedly written by females. The researchers attached a photograph to each form that was previously rated for attractiveness (attractive, unattractive). In addition, they created a cover letter stating that UCLA was attempting to streamline its student mental health intake procedures, and needed therapists to help by evaluating a case for symptom severity and the potential for therapeutic success. Packets consisting of the cover letter and one of the four cases were sent to a randomly selected, nationwide sample of psychiatrists, psychologists, and social workers. Murray and Abramson found that the "attractive" cases were judged to have less severe symptoms, and that the therapists more willing to work with these cases and more confident of their ability to be successful with them.
Obviously, there are many more ways to apply the study of human sexuality to clinical issues. Recently, one important focus has been on risky sex and psychotherapy for HIV-infected individuals (e.g., Farber & Schwartz, 1997; Fontaine, 1995; Gunther, Crandels, Williams, & Swain, 1998; Henry, 1996; Kelly, 1998; Markowitz et al., 1998; Markowitz, Rabkin, & Perry, 1994; Millan & Caban, 1996; Sarwer & Crawford, 1994).PROFESSIONAL DEVELOPMENT
Before we conclude our overview of research on human sexuality, we would like to mention some of the professional societies and journals that support and publish the kind of sex research that we have discussed. The principal focus of our survey is sex research, broadly defined, and on clinical applications such as sex education and therapy. Given our focus, we do not cover several sex-related disciplines (e.g., public health, HIV/ AIDS, genitourinary medicine), each of which has professional organizations and publications that interest sex researchers.
Founded in 1957, the Society for the Scientific Study of Sexuality (SSSS) is the oldest organization in the United States that is devoted to the professional study of human sexuality. With over 1,100 members, the Society is also one of the largest sex research organizations in the world. According to the Society's literature, "SSSS brings together an international group of professionals who believe in the importance of both the production of quality research and the clinical, educational, and social applications of research related to all aspects of sexuality." Each year, the Society holds a national conference and three smaller regional conferences.
The Society's official publication, The Journal of Sex Research (JSR), is arguably the nation's premier sex research journal. Published quarterly, JSR is "designed to stimulate research and to promote an interdisciplinary understanding of the diverse topic in contemporary sexual science." In addition to original reports of empirical research, the journal publishes theoretical essays, literature reviews, methodological notes, historical articles, clinical reports, teaching papers, book reviews, and letters to the editor. The Society also publishes the Annual Review of Sex Research.
Sex education and related topics are the focus of the Sexuality Information and Education Council of the United States (SIECUS). SIECUS is a national, nonprofit organization that was founded in 1964 to develop, collect, and disseminate information about human sexuality; to promote comprehensive sexuality education; and to advocate the right of individuals to make responsible sexual choices. SIECUS maintains a comprehensive sexuality library and publishes numerous pamphlets, booklets, and bibliographies for professionals and the general public alike. Their bimonthly journal, SIECUS Report, offers "ground-breaking articles written by prominent leaders . . . updates on relevant advocacy and legislative issues, timely reviews of newly released books and videos, and announcements for conferences and meetings." SIECUS's sister organization in Canada is called the Sex Information & Education Council of Canada (SIECCAN). SIECCAN publishes the quarterly journal The Canadian Journal of Human Sexuality, which covers a range of topics of interest to sexuality researchers.
The American Association of Sex Educators, Counselors, and Therapists (AASECT) was founded in 1967. AASECT is devoted to "promoting under-standing of human sexuality and healthy sexual behavior." The association "certifies qualified health and mental health practitioners in dealing expertly and ethically with the sexuality concerns of individuals and couples" and offers a broad range of education and training activities. In addition, AASECT publishes The Journal of Sex Education and Therapy.
The International Academy of Sex Research has some 200 members. The Academy's official publication is the Archives of Sexual Behavior. This British journal is published bimonthly and "reports the latest research trends in the science of human sexual behavior, bringing together high-quality submissions from such diverse fields as psychology, psychiatry, biology, ethology, endocrinology, and sociology."
In addition to the aforementioned journals, individuals and researchers interested in human sexuality have access to journals that address different areas. For example, the quarterly Journal of the History of Sexuality "illuminate[s] the history of sexuality in all its expressions, recognizing various differences of class, culture, gender, race, and sexual orientation." Another quarterly, Sexualities: Studies in Culture and Society, publishes "articles, reviews and scholarly comment on the shifting nature of human sexualities." The similarly titled Sexuality & Culture provides a forum for "the discussion and analysis of ethical, cultural, psychological, social, and political issues related to sexual relationships and sexual behavior." The 1997 and 1998 volumes focused, respectively, on sexual harassment/ sexual consent and sex work. Another quarterly, the Journal of Psychology & Human Sexuality, covers a broad range of sex-related topics from a variety of perspectives, including clinical, counseling, educational, social, experimental, psychoendocrinological, and psychoneurological research.
There are several special-interest journals related to sexuality. Gender, sex roles, and sexual orientation are the subject of GLQ: A Journal of Lesbian and Gay Studies; Journal of Gay, Lesbian, and Bisexual Identity; Journal of Homosexuality; and Sex Roles: A Journal of Research. Clinical and therapeutic issues are discussed in the Journal of Gay and Lesbian Psychotherapy; Journal of Sex and Marital Therapy; Journal of Sex Education and Therapy; Journal of Social Work and Human Sexuality; International Journal of Impotence Research; Medical Aspects of Human Sexuality; Sexual Abuse: A Journal of Research and Treatment (formerly, Annals of Sex Research); Sexual Addiction & Compulsivity; and Sexuality and Disability.
Having a number of professional organizations, each with its own unique focus, and having access to multiple avenues for communicating with other professionals are indeed blessings. However, to ensure that future researchers experience the same bounty, those of us who are currently active will need to become more involved in shaping and supporting the formal sex education that children, adolescents, and young adults receive at home, in school, and from health care and social service providers. Some of these individuals will participate directly in studies of human sexuality; almost all will be asked to support the funding of such studies. As we stated earlier, sexual literacy is an important societal goal that benefits the individual, the researcher, and society.
CONCLUSION
Sex and sexuality are important concerns of laypersons and sex researchers-- and rightly so. The behaviors, the practices and patterns, and the thoughts and feelings reflected in the phrase "sex and sexuality" are fundamentally important to our individual and collective health and well-being, to our sense of who we are and our place in the world. Thus, whereas many Americans have been able to abandon social distinctions (and the inherent privileges that accompany such distinctions) based on social class, country of origin, and race, a good number are reluctant to abandon social distinctions based on gender and gender roles.
Research on human sexuality can play an important role in the ongoing transition. In particular, sex researchers can obtain accurate and timely information on the determinants of and mechanisms regulating all manner of sexual attraction and expression; on what it means to be male or female (straight, gay, lesbian, bisexual, or transsexual) in a particular culture; and on the individual and social consequences of sexual behavior. Because sex researchers continue to pay close attention to issues such as design (and, specifically, the design of the study, the items and instruments, and the sample) and data collection, their methods are often more persuasive to the lay public than other modes of inquiry and validation.
In this chapter, we have endeavored to show that sex research has benefited greatly from numerous insights and contributions from the life sciences (e.g., evolutionary psychology and biologically based theories of human sexual behavior), the humanities (e.g., the theories founded on postmodernism, deconstruction, and feminism), the cognitive sciences (e.g., script theory), mathematics (e.g., epidemiological modeling), and prevention research (with its theories and models of behavior change). We firmly believe that future contributions from these disciplines will continue to provide some of the impetus for the important work done by clinicians and other service providers.
However, we encourage sex researchers to consider more integrative theories of human sexual behavior. To date, the use of specific models of sexual behavior, rigorous methodologies, and sophisticated statistical techniques have not yielded satisfactory answers to important questions such as Why does sex feel good? or Why are we sexually attracted to given individuals (or objects of desire, as the case may be)? Clearly, more informative, comprehensive theories will be of greater use to clinicians working with individuals who are struggling with issues of sexual identity, as well as prevention researchers who often ask participants to enact behaviors that individuals find difficult to initiate and maintain (e.g., use condoms to avoid sexually transmitted infections).