Dangerous data: Seeing social surveys through the sexuality prism

Abstract

Social surveys both reflect and shape beliefs about sexuality. Social norms construct the “authorized vocabulary” of surveys and the resulting data influence the research questions that can be answered and the policies likely to be inspired by study findings. Scholars have called for balancing attention to pleasure vs. danger and normative vs. non-normative practices in studies of sexuality as well as for collection of data on sexual desires, behaviors, and identities. We combine these calls into what we term the sexuality prism. To better understand how data about sexuality are typically collected and what research they facilitate or constrain, we analyze six decades of materials from four of the largest social surveys in the United States and five national surveys focused on sexuality, health, and family formation. We find that these surveys do not allow for investigations of the full sexuality prism. Instead, they tend to assume and narrowly investigate the “charmed circle” of sexuality: heterosexual, married, monogamous, and potentially procreative couplings. When surveys ask about non-normative practices, they do so primarily in the context of risk (e.g. sexually transmitted diseases) and ignore non-normative practices that are not deemed “risky.” The focus on risk likely explains the greater attention to sexual behaviors and the shortage of questions about sexual desires and identities. Moreover, most questions about sexual practices highlight the dangers of sex, rather than the pleasures. Not only does this severely limit the scope of U.S. sexuality research, it also means that, individually and collectively, these surveys reify “sex negativity.”

Keywords

Social surveys sexual desires sexual behaviors sexual identities pleasure

The time has come to think about surveying sexuality. In an era when funding for sexuality research in the USA is limited, available survey data shape both the research questions that can be asked and the answers that can be found. The resulting scholarship can have consequences far beyond academia. By influencing knowledge, survey research also shapes social beliefs, practices, and policies. Population science on sexuality drives federal policy and research priorities (Riley and McCarthy, 2003) and figures centrally in the use of social science expertise in legal arenas (Biblarz and Stacey, 2010; Stacey, 2004). Moreover, although surveys are often seen as simply collecting information, survey questions also teach. When respondents read survey questions, they receive lessons in social norms, as does anyone who utilizes the data or engages with survey findings in the media (Igo, 2008; Lee et al., 2014; Norton, 2013; Nowakowski et al., 2016).

Despite its influence, many sexuality and gender scholars dismiss survey research as part of broader queer and feminist critiques of quantitative methods (Baumle, 2018; Undurraga, 2010; Williams, 2006). Quantitative methods have long been characterized as both positivist and reductionist, and thus unable to represent systems of oppression such as sexuality, gender, and race with complexity and nuance. However, surveys can and should be “rehabilitated” (Oakley, 1998). Therefore, we aim to contribute to the growing “sociology of quantification” (Espeland and Stevens, 2008) by asking what surveys measure and how they do so.

Previous studies of sexuality in surveys tend to offer a deep analysis of a single survey or a single measure or concept (e.g. sexual orientation). Much of the existing literature has focused on the measurement and classification of non-heterosexuality in particular (Baumle et al., 2009; Cruz, 2017; Michaels and Lhomond, 2006; Moore and Stambolis-Ruhstorfer, 2013; Powell et al., 2010; Sumerau et al., 2017). This scholarship has worked to push measurement beyond heteronormative perspectives, expanding understanding of the sexuality system.

To extend this important work, we take a broader approach and examine how nine U.S. national surveys spanning six decades of data collection mention and measure sexuality in all of its forms. Sexuality scholars have long argued that both academic research and broader cultural understandings have focused on normative sexual practices as well as the dangers of sexuality, to the detriment of pleasures and stigmatized practices (e.g. Ford et al., 2019; Jones, 2019; Rubin, 1984; Vance, 1984). Sexuality scholars have also urged researchers to attend to, and differentiate between, sexual desires, behaviors, and identities (e.g. Diamond, 2008; Laumann et al., 1994a; Ward, 2015). Together, these calls represent a multidimensional perspective that we term the sexuality prism. This conceptual approach attends to four main facets of sexuality: sexual desires, identities, behaviors, and outcomes. It also encompasses a range within each facet, including both normative and non-normative practices as well as their dangerous and pleasurable consequences. We analyze the survey questionnaires, user guides, and other technical materials with the organizing principle of the sexuality prism in mind.

Our analysis reveals that these nine surveys tend to treat relationship status as a proxy for sexual behaviors rather than asking directly about respondents’ sexuality. Relationship status questions overwhelmingly focus on marriage, monogamy, and heterosexuality, ignoring other relationship forms. When they do ask about sexuality directly, these surveys generally pose questions about sexual behaviors in the context of “risk,” including pregnancy, sexually transmitted diseases (STDs), and sexual assault. This line of questioning sometimes acknowledges non-normative behaviors but also explicitly and implicitly links sexuality and danger, reinforcing “sex negativity” (Rubin, 1984). Moreover, it overlooks other key aspects of the sexuality prism, including positive outcomes of sexuality. As a result, there is little data available for researchers to study desires, identities, pleasures, and behaviors deemed less “risky.” Together, these patterns flatten available images of sexuality and reinforce normative assumptions about sexuality as dangerous.

Measuring and conceptualizing sexuality

Systems of official classification, such as those used in large-scale surveys, both reflect and construct social life (Bowker and Star, 2000; Law, 2009); they are products of historical context that tend to reproduce dominant norms, making them difficult to change. Foucault (1978) argued that ways of speaking about sexuality follow “rules of propriety” dictating how, where, when, and with whom conversations could occur. This “authorized vocabulary” of sexuality developed by the 19th century, alongside a modern system of governance that aimed to discipline life through the exercise of “biopower.”

The origins of scientific studies of sexuality are entwined with the growth of population statistics. The state’s interest in regulating marriage, family, childbearing, and physical health through monitoring population “state-istics” was the precursor to the modern field of demography (Cohen, 1999; Louckx and Vanderstraeten, 2015). Surveys became the authoritative source for population-level information on these topics beginning in the mid-20th century, fueled by the development of advanced methods for sampling and statistical analysis (Igo, 2008).

Historically, questions about sexuality were often included in surveys in response to moral panics or perceived challenges to dominant norms (Bullough, 1994; Parker and Gagnon, 1994). As Ericksen and Steffen (1999) detail, the earliest American surveys that included questions about sexuality surveyed young men in the military and universities, and focused on issues of “social hygiene” such as masturbation, venereal disease, and paid sex. The first surveys to ask women about sexuality were conducted at women’s colleges and included questions about masturbation, monthly cycles of sexual desire, and lesbian sex. As concerns about declines in American fertility grew during the 1930s, larger surveys conducted by demographers began to ask questions about sexuality in the context of contraception, family formation, and birth outcomes. By the 1970s, the availability of contraception and further reduced fertility among married white women shifted attention to the fertility of Black women and teenagers, which helped create “the social problem of adolescent sexuality and teen pregnancy” (Ericksen and Steffen, 1999: 90). More recently, the HIV/AIDS crisis led to the adoption of questions on sexuality across more surveys and expanded the topics covered (Baumle, 2013).

The contemporary history of sexuality in surveys also reflects the choices of, and constraints faced by, social science researchers generally and survey administrators specifically. Question wording and survey instructions reflect a combination of researcher conceptualizations of sexuality and general social norms, with large-scale surveys in particular facing external pressures related to funding and collaboration with the federal government (Miller, 1995; Shiode, 2016). The storied history of the National Health and Social Life Survey (NHSLS), in which a team of social scientists lost their federal funding amid a media and political frenzy for daring to include questions about non-procreative practices in their survey of sexuality (Laumann et al., 1994b), exemplifies the authorized vocabulary of sexuality and its limits on knowledge production. It also served as a cautionary tale for the social scientists behind the contemporary National Longitudinal Study of Adolescent to Adult Health (Add Health) whose successful bid for federal funding rode on the positioning of a survey of teenage sexuality unequivocally within a health frame (Shiode, 2016). However, even as the increasing rhetorical pairing of sexuality and health can “sanitize” the perceived dirtiness of sex, it risks reinforcing the idea that sexuality is only a legitimate object of study when it is linked to health (Epstein and Mamo, 2017).

Calls for change

Over the last several decades there have been numerous calls to expand the study of sexuality beyond the narrow confines of the current authorized vocabulary shaped by moral panics. We focus on three here: (1) highlighting the pleasures of sexuality; (2) attending to the “outer limits” of sexual practices; and (3) measuring desires, behaviors, and identities. We combine these calls for change into a single conceptual framework, the sexuality prism, which highlights four main facets of sexuality: desires, behaviors, identities, and outcomes.

One foundational call came from Carole Vance (1984), who criticized the activist and academic focus in the United States on the dangers of sexuality and the corresponding lack of attention to sexual pleasures. Scholars continue to critique this tendency to focus on danger, demonstrating that the lack of attention to pleasure is detrimental to both sexuality scholarship and everyday experiences of sexuality (Jones, 2019; Schalet, 2011; Tepper, 2000), and efforts to recognize sexual pleasure as “a component of holistic health and well-being” (WASH, 2008) are growing around the world. Recent scholarship also demonstrates the benefits of centering pleasure in analysis, including identifying how pleasure shapes sexual behaviors (e.g. Burke, 2016; Jones, 2020; Miller-Young, 2014; Nash, 2014; Orne, 2017). However, the relative dearth of research in this area has consequences for clinical practice and public health policy (Ford et al., 2019) as well as sex-positive understandings of human development (Harden, 2014).

In tandem with Vance’s call to include pleasure, Gayle Rubin argued that “the time has come to think about sex” (1984: 143) and challenged the system of “sexual stratification” in which certain sexual behaviors are deemed valuable and worthy of attention whereas others are marginalized or silenced. Rubin described a “charmed circle” of sexuality that is seen as “good, normal, natural, blessed” (married, monogamous, procreative, vanilla, private) positioned opposite the “outer limits” of sexuality (homosexual, promiscuous, commercial, pornographic, sadomasochistic) which are “bad, abnormal, unnatural, damned” (153). Rubin called for a politics of sexuality that takes seriously all sexual practices and the consequences of their place in the “hierarchy of sexual value” (152).

More recently, scholars have argued that research should attend to multiple dimensions of sexuality, specifically desires, behaviors, and identities (Mishel, 2019; Savin-Williams and Ream, 2007; Silva, 2019; SMART, 2009; van Anders, 2015). Although these three dimensions are often thought to “match,” that is not always the case. A person’s sexual desires (i.e. attractions, fantasies, or motivations for sex) are not always translated into behaviors (i.e. sexual actions). Similarly, a person’s sexual identity (i.e. personal and communal understandings about oneself, often codified into categories such as gay, straight, etc.) is not a perfect indicator of their sexual desires and behaviors, and vice versa (e.g. Ward, 2015). Thus, measuring just one dimension provides incomplete information; understanding human sexuality requires attention to all three aspects.

It is critical that sexuality scholars attend to surveys—and that survey research attends to sexuality scholarship—so that the resulting data reflect the best practices of the sexualities field. To that end, we combine these three calls for change in sexuality scholarship into a single organizing concept: the sexuality prism (see Figure 1). In this article, we ask whether surveys attend to all four facets of sexuality by measuring desires, behaviors, identities, and their “outcomes” (i.e. the results of sexuality, including procreation, STDs, and pleasure and other emotions).¹ Within each facet, we also explore whether the full range of sexuality is represented, including both charmed circle and outer limits desires, behaviors, and identities as well as a balance of outcomes between pleasure and danger. In our evaluation of whether these surveys facilitate scholarship on sexuality’s multiple dimensions, our analysis reveals what the authorized vocabulary of sexuality has been over the past six decades of U.S. survey research.

Figure 1.

The sexuality prism. Note: This depiction is intended to evoke looking through one end of a rectangular prism so as to see all four facets at once. The shading is included to help the image look three-dimensional and darker shading reflects more frequent measurement in the surveys we examined. The orientation of the prism is not intended to suggest a hierarchy of facets; rotating the prism in various directions may help to highlight different connections between the facets for different purposes. The four aspects of sexuality are depicted as facets so as to encourage examination of a range within each.

Methods

We explore these questions through an examination of four of the largest and longest-running surveys in the USA, as well as five U.S. specialty surveys focused on sexuality, health, and family formation. The four major social surveys are the American National Election Study (ANES), Panel Study of Income Dynamics (PSID), General Social Survey (GSS), and National Longitudinal Study of Youth (NLSY). The five specialty surveys are the National Health Interview Survey (NHIS), National Survey of Family Growth (NSFG), National Health and Social Life Survey (NHSLS), National Longitudinal Study of Adolescent to Adult Health (Add Health), and National Sexual Health Survey (NSHS). See Table 1 for more details on each survey. We chose a range of surveys in order to allow for a variety of approaches to studying and measuring sexuality. Sexuality specific surveys such as the NHSLS and NSHS were selected to represent what might be considered “best practices” on these topics, and broader surveys of health and family formation were included to represent the traditional authorized vocabulary for studies of sexuality. We also incorporated the general surveys because of their long history as a central source of data in social science research and teaching, with important implications for academic knowledge, cultural beliefs, and public policy.

Table 1.

Description of survey sample.

	General surveys				Specialty surveys
	American National Election Study	Panel Study of Income Dynamics	General Social Survey	National Longitudinal Survey of Youth	National Health Interview Survey	National Survey of Family Growth	National Health and Social Life Survey	National Longitudinal Study of Adolescent to Adult Health	National Sexual Health Survey
Years active	1948–present	1968–present	1972–present	1979–present	1957–present	1973–present	1992	1994–present	1996
Survey design	Cross-sectional, pre-/post-election interviews; occasional multi-year panel studies	Longitudinal	Cross-sectional; three-wave panel studies 2006–2014	Longitudinal	Cross-sectional, continuous sampling	Cross-sectional; continuous sample began in 2006; just females until 2002	One-time specialty survey	Longitudinal	One-time specialty survey
Sample size	About 1,500 eligible voters	Began with 4,800 households	About 1,500 adults in households	Began with 12,686 14–22-year-olds	More than 100,000 people annually, from 35,000–40,000 households	Varies; 2006–2010: 22,682, ages 15–44	3,432 adults, ages 18–59.	In-school sample: 90,118 students; in-home interviews began with 20,745 7th–12th graders	8,467 adults in households in the continental U.S.
Funders^a	NSF	OEO NSF NIA NICHD	NSF	NICHD	DHHS CDC NCHS	DHHS CDC NCHS	Private foundations^b	NICHD	NIMH

^aFunders are listed with acronyms for ease of comparison. CDC: Centers for Disease Control; DHHS: Department of Health and Human Services; NCHS: National Center for Health Statistics; NIA: National Institute on Aging; NICHD: Eunice Kennedy Shriver National Institute of Child Health and Human Development; NIMH: National Institute of Mental Health; NSF: National Science Foundation; OEO: Office of Economic Opportunity of the United States Department of Commerce.

^bThe private foundations were The Foundation for AIDS Research, Robert Wood Johnson Foundation, Henry J. Kaiser Family Foundation, Rockefeller Foundation, Andrew Mellon Foundation, The John D. and Catherin T. MacArthur Foundation and New York Community Trust.

We collected all of the questionnaires, codebooks, interviewer instructions, and user guides produced for each survey and made publicly available online for every round of interviews from their inception through 2011. The collected documents cover 190 survey-years and a total of 119,105 pages, an average of about 600 pages per survey-year. To explore the authorized vocabulary of sexuality, each document was examined in its entirety for passages and questions of interest. To ensure that no relevant sections were missed, we searched each document for 62 keywords ranging from “sex” to “married” to “pleasure.” Text was extracted for qualitative coding (in one or more pages) when it related to sexuality in any way, including relationship status, opinion or knowledge questions, or any of the sexuality prism’s four facets. Extracts included moments where surveys simply mentioned sexuality, such as in interviewer instructions or questions about attitudes or beliefs, as well as when they directly measured sexuality, such as asking respondents explicitly about desires, behaviors, identities, or outcomes.

To capture trends in how each survey conceptualized sexuality, we extracted one example of each type of sexuality related question, instruction to interviewers, and other documentation in a given survey-year. For example, in 1993, the NLSY included more than 300 questions about the respondent’s “husband/wife” and we extracted the first example for coding because the number of categories and the criteria for membership in those categories were the same for each question. If, in the same year, a survey also used a different conceptualization, such as referencing the respondent’s “husband/wife/partner,” we also collected that example to capture the range of conceptualizations that existed in the survey-year. Following the parameters described above, we collected and analyzed 3434 excerpts related to sexuality.

This study combines techniques from textual and discourse analysis with those from traditional content analysis (Mills, 1997; Neuendorf, 2002; Titscher et al., 2000). Each question, variable description, or text excerpt was coded using ATLAS.ti, a qualitative data analysis software program ideal for handling projects with a large number of primary documents. Codes were developed inductively to respond to patterns that emerged from the data. We generated a total of 263 codes clustered into 16 questions, including: what topics were being queried (e.g. condom usage, orgasms, or pregnancy), what types of relationships were mentioned (e.g. marriage, cohabitating, or a non-consensual interaction), and what sexual acts were mentioned (e.g. vaginal, oral, or masturbation). We used these codes to analyze the methodological decisions made by the nine surveys as to which aspects of sexuality to measure, how to measure them, and which forms of sexual practices and relationships to acknowledge.

To our knowledge, this is the most comprehensive study of sexuality content in social surveys to date. As noted above, most previous studies focus on a single concept, such as sexual orientation, or a single data source, such as the U.S. Census. Moreover, broader narrative histories of sexuality in surveys (e.g. Ericksen and Steffen, 1999) do not examine the actual questionnaires and other technical documentation. We address these gaps by undertaking a rigorous and theoretically motivated analysis of question wording, question order, response options, and interviewer instructions across a range of surveys.

Of course, data collection for many of these surveys is ongoing, and a limitation of our full analysis is that it ends in 2011. Since then, there has been considerable academic and activist attention to sexuality in surveys (Baumle, 2018; Miller and Ryan, 2011; SOGI-IWG, 2016), and notable federal efforts to define best practices for SOGI (sexual orientation and gender identity) measurement (see SOGI-IWG, 2016). However, to date, smaller-scale research using non-representative surveys and continued calls for better data have largely outpaced adding new measures to large-scale surveys and federally funded data collection (NASEM, 2020; see also Wang, 2017). Supplementary analysis of available materials for the 2011–2019 period from all four of the general surveys and the three specialty surveys that include repeated data collection did not reveal evidence that runs counter to the overall patterns highlighted below. To mirror the first step of our original analysis, we downloaded 18,746 pages of questionnaires and codebooks from post-2011 survey waves and executed comprehensive keyword searches using MAXQDA qualitative analysis software to identify excerpts that either mentioned or measured sexuality. Keyword searches of these seven surveys revealed that in the post-2011 period these surveys continued to focus on relationship status, “risky” behaviors, and negative outcomes, to the detriment of measurement of desires, less “risky” behaviors, and positive outcomes. Moreover, only NHIS added a question on sexual orientation (in 2013). Of the other specialty surveys, sexual orientation questions were included in the NHSLS (1992) and NSHS (1996) and were added to Add Health in 2001 and the NSFG in 2002. Of the general surveys, both the GSS and ANES added sexual orientation questions in 2008, whereas the PSID and NSLY are yet to include SOGI questions. The relatively slow pace of change in terms of including measures of sexual identity, despite increasing public interest and academic pressure, also fits our conclusions about broader trends in these surveys over the past six decades.

Although our analysis of national surveys in the United States is extensive, it is not exhaustive. We do not cover all types of surveys that measure sexuality and our data do not speak directly to how the surveys were produced or received. Our focus was on large-scale surveys with publicly available materials because they are the most likely to be used in research and teaching and cited in the media. This decision likely overrepresents federally funded efforts that mandate data sharing so our findings may not generalize to smaller U.S. surveys or industry-funded research, such as the National Survey of Sexual Health and Health Behavior. Surveys from other countries likely would reveal other patterns, as well, depending on whether sex negativity is more or less prominent in the given society (for an overview of international surveys with relevant sexuality measures, see Savin-Williams and Vrangalova, 2013). Comparative studies both within and across countries that explicitly examine the roles of both funding and cultural beliefs in shaping survey production, as well as how the beliefs of respondents, researchers, and others are affected by the conceptualization and measurements of sexuality in surveys would be fruitful areas of future research.

Surveying sexuality

Despite many calls over the past several decades to broaden studies of sexuality, the surveys we examined continue to focus on one facet of the sexuality prism: behaviors perceived as putting respondents at risk for the outcomes of pregnancy and STDs. Although, the perceived risks changed over six decades, the overall focus on sexual behavior did not. Attention to pleasure, desires, and identities remained rare throughout this time period. By contrast, queries about relationship status, which function as proxies for sexual activity, were nearly universal. These indirect questions consistently prioritized the charmed circle of marriage, monogamy, and heterosexuality. Direct questions about sexual behaviors were much less common and generally occurred in the context of risk or other negative aspects of sexuality such as assault. Sexual stratification was evident here as well: in assessing respondents’ risk for procreation, questions focused on potentially reproductive heterosexual sex whereas questions assessing risk for STDs acknowledged oral, anal, and same-sex sex.

Notably, both the general and specialty surveys adopted similar approaches. Although the specialty surveys were more likely to include questions about desires, identities, and pleasures, such inclusion was not universal and all nine surveys highlighted risky behaviors and the dangers of sex. The similarity of our findings across the surveys is striking, particularly given variation in their formats and funders (see Table 1). Thus, we organize our presentation of results below by theme rather than survey or time period, highlighting key exceptions or outliers over time and between surveys as needed.

Centering the “charmed circle”

Sexuality most frequently appeared in these surveys in the guise of relationship status. As demonstrated in Table 2, compared to measurement of desires, behaviors, and identities, questions about relationship status were much more common. Of the 190 survey-years analyzed, 185 (97%) contained at least one direct question about the respondent’s relationship status. The five that did not come from just two surveys (ANES and NHIS) and researchers using that data could obtain relationship status for those respondents via other means, such as the separate household roster. Relationship status could be considered an indicator of the potential for normative sexual behavior (e.g. state-sanctioned, monogamous, procreative sex). However, relationship status is a poor proxy for sexual practices and using it as such ignores Rubin’s (1984) call to attend to sexual practices outside of the “charmed circle.”

Table 2.

Number of survey-years, pages, and excerpts and percent of survey-years measuring key aspects of sexuality.

	General surveys				Specialty surveys					Total
	American National Election Study	Panel Study of Income Dynamics	General Social Survey	National Longitudinal Survey of Youth	National Health Interview Survey	National Survey of Family Growth	National Health and Social Life Survey	National Longitudinal Study of Adolescent to Adult Health	National Sexual Health Survey	9
Survey-years^a	39	37	28	24	49^b	7 “cycles”	1	4	1	190
Total pages^a	35,470	17,539	8,689	8,462	17,784	20,788	670	3,587	6,116	119,105
Sexuality Excerpts Per Year (Total)	13.3 (520)	13.6 (503)	23.8 (666)	10.7 (256)	5.4 (267)	84.3 (590)	200 (200)	69.8 (279)	153 (153)	Average: 67.8 Total: 3,434
Relationship status^c	92%	100%	100%	100%	96%	100%	100%	100%	100%	97%
Desires^c	0	0	4%	0	0	44%	100%	100%	0	4%
Behaviors^c	0	3%	39%	17%	43%	100%	100%	100%	100%	26%
Identities^c	3%	0	7%	0	0	44%	100%	50%	100%	5%

^aCounts for survey years and the total pages examined reflect data collection through 2011.

^bAnnual questionnaires for NHIS are only publicly available starting in 1962.

^cRelationship status, desires, behaviors, and identities are percentage of survey-years with questions that measure these factors. Questions or text that mention them, but do not measure them, are not included. Only survey-years with direct questions are included. Survey-years where the information is carried forward are not included, as all four aspects of sexuality can change from year to year.

There are three main patterns to how relationships were measured across these surveys: (1) they were usually defined in reference to marriage; (2) it was most common to ask about one partner rather than multiple partners; and (3) heterosexual couplings were acknowledged more than alternatives. Together, these patterns describe the center of Rubin’s charmed circle: married, monogamous, and heterosexual.

Surveys focus on marriage by defining each relationship response option relative to marriage. Standard answer options included “married” and variations on “formerly married” (“widowed,” “separated,” “divorced”) and “never” or “not (yet) married” (“single”). This question from the NSLY is typical:

(NLSY, 2010: 8)

When surveys do acknowledge non-marital relationships, this recognition still signals that marriage is the norm to which everything else is compared. For example, over time, every survey we examined added a “partner” answer option (see also Westbrook and Saperstein, 2015). However, when included, “partner” was usually asked in a separate question or listed after marital relationships, as in this item from the ANES:

(ANES, 2002: 563)

Note that “partner” appeared even after “never married” on the list. The “{VOL}” indicates that it was not an answer option offered to the respondent; it was only recorded if the respondent volunteered the information. “Partner” is devalued when it appears separately, last in the list of response options, or is not an offered answer option.

The surveys we examined also focused on monogamy. Nearly all survey-years (182 of 190, or 96%) included question formats that counted only one sexual partner. For example, Add Health acknowledged that a respondent may have multiple sexual partners, but instructed interviewers to only record data about one:

(Add Health, 2007–2008: 325)

Data collected following these instructions paints only a partial picture of respondents' sex lives. About half of survey-years included at least one question or response option that acknowledged the possibility of multiple, simultaneous sexual partners (109 of 190, or 57%). However, these questions were separate from relationship status questions and often implied that the non-monogamy was non-consensual (i.e. “cheating”), such as this question from the GSS with an abbreviation for “ever stray” as the variable name:

(GSS, 2010: 225)

Such questions offer data on the prevalence of non-monogamy, but because they do not distinguish between consensual and non-consensual forms, they effectively erase polyamory and other consensual forms of non-monogamy. In doing so, they both reflect and reify “compulsory monogamy” and “mononormativity” (Emens, 2004; Pieper and Bauer, 2005).

In addition to a focus on marriage and monogamy, these surveys focus heavily on heterosexuality. A majority of the survey-years analyzed (128 of 190, or 67%) included questions that explicitly recognized only an opposite-sex partner, such as this question from the 2007 NSFG:

(NSFG Male Questionnaire, 2007: 5)

In comparison, a minority of survey-years included questions that explicitly included both same- and opposite-sex partners (63 of 190, or 33%). Often the inclusion of non-heterosexual relationships, like non-marital relationships and non-monogamy, reproduced their outer limits status, as in this excerpt from the GSS:

(GSS, 1996: 70)

In sum, counting and describing respondents’ relationship status appears to be central to the aims of the social surveys we analyzed, but the surveys did so in ways that disproportionately focus on the charmed circle of marriage, monogamy, and heterosexuality. When alternatives appeared, they were typically devalued and marked as other. This has consequences not only for what can be learned from survey data but also in terms of how survey practices shape cultural beliefs: focusing on normatively approved relationships and disregarding outer limits alternatives reinforces their stigmatized status.

Discounting desires and identities

Although scholars have argued that surveys should ask questions about a combination of desires, behaviors, and identities, most of the surveys we examined did not do so. As illustrated in Table 2, desire, in particular, was almost never mentioned in these surveys, let alone measured. Of the nine surveys we examined, only four ever included any mention of sexual desire. Of those, three (Add Health, GSS, NSFG) focused their questioning on which sexes or genders a respondent is “attracted” to. This question from the NSFG is typical:

(NFSG Female Questionnaire, 2008: 356)

Further, in a theme that runs throughout our analysis, we find that even questions about desire were asked in the context of risk. For example, the 2008 NSFG question appeared in a section of the questionnaire titled “Sexual Attraction, Orientation, & Experience with STDs.”

Though sexual object is a key aspect of desire, it is not the only one. In addition to with whom you may want to be sexual there are also sexual “aims,” meaning activities in which you would like to engage (Freud, 1949). Of the nine surveys we examined, only the NHSLS measured fantasies or actions the respondent finds desirable, as in this question:

(NHSLS, 1992: 591)

Unlike most surveys, the NHSLS allowed researchers to examine how appealing the respondent finds a wide variety of behaviors from anal sex, to using dildos, to watching their partner strip. Counter to the dominant trend across surveys, several of the behaviors mentioned in NHSLS questions about desire involve low risk for STD transmission. By attending to desires and asking about low risk activities, these types of questions facilitate research on sexuality that is not focused on danger. They also allow researchers who are interested in negative outcomes of sex to better understand those, as desires motivate “risky” behaviors, and researchers concerned about risk should also attend to low risk activities.

Like questions about desires, questions about sexual identities were rare in these surveys (Table 2). Survey research has acknowledged sexual orientation in questions about personal opinions and public attitudes since the 1970s. Surveys have also asked about same-sex sexual behaviors with some frequency since the late 1980s, yet fewer than 5% of the survey-years we examined included a direct measure of sexual identity. Although NHSLS and NSHS (fielded in 1992 and 1996, respectively) included sexual identity questions, the general surveys did not add them until the late 2000s, nearly 20 years after they started asking about same-sex behaviors. Notably, unlike identity questions around gender that only allow for binary options (i.e. male and female) and see changes in gender identity over time as errors (Westbrook and Saperstein, 2015), these surveys did not “foreclose fluidity” (Sumerau et al., 2020) around sexuality. Instead, when asking about sexual identity, they always included at least three answer options, such as this question from the GSS:

(GSS, 2008: 237)

Moreover, they did not treat changes in sexual identity over time as errors. The addition of these identity measures reflects the growing inclusion of gay men, lesbians, and bisexuals in civic life over the last three decades (Baumle et al., 2009; Gates, 2011) as well as the official public health declaration that LGBTQ people are a “population at risk” (NIMHD, 2016). Like outer limits sexual behaviors, we find questions about identity are often confined to sections of the questionnaire reserved for “risky” or “deviant” practices, rather than appearing with the other identity questions such as race, ethnicity, religion, or political affiliation. This reifies sex negativity and the stigma against non-heterosexual identities. Although attention to sexual identity has increased over time, the limited measurement of desires and identities overall hinders researchers’ ability to study all facets of the sexuality prism, including exploring when desires, behaviors, and identities “match” or diverge, as well as the relationships between desires, identities, and sexual outcomes.

Recognition through “risk”

Compared to desires and identities, direct questions about sexual behaviors appeared with some frequency. About 26% of the survey-years analyzed included items measuring sexual behaviors. However, such questions were almost always asked in the context of risk, including risk for pregnancy, STDs, and sexual assault. When attending to the risk of pregnancy, the surveys we examined focused exclusively on heterosexual, potentially procreative sex. By contrast, when concentrating on STD risk, these surveys also measured sexual behaviors that cannot result in pregnancy. The attention to risk allows for the recognition of outer limits sexual behaviors as well as experiences often shrouded in shame, such as sexual assault. At the same time, it limits sexuality research by focusing only on risky behaviors and negative outcomes, rather than pleasure and behaviors deemed less risky.

Preoccupation with procreation

In the surveys we examined, interest in procreation was often the main motivation for asking respondents about sexual behaviors. In the specialty surveys in particular, respondents were regularly asked if they were heterosexually active, menstruating, and/or using contraception. Ascertaining whether someone is at risk of pregnancy was so central that they often asked respondents in multiple ways and instructed interviewers to ensure that respondents answered these questions. For example, in a set of questions asked of females² about their first “sexual intercourse with a man,” the NSFG tells interviewers:

(NSFG, 1995: 7583)

Those whose answers indicated that they were “at risk” of pregnancy were asked many follow-up questions whereas those who were not were routed past them. Interest in potential procreation also opened the door to asking about behaviors outside of the charmed circle such as sex outside of marriage or at the “wrong” age. However, because pregnancy was the dominant outcome of interest, the measured behaviors relate to potentially procreative (i.e. heterosexual) sex and the relationship status in which such sex takes place. Behaviors that cannot result in pregnancy, such as same-sex sex, were asked about much less often; they literally do not “count” when surveys are focused exclusively on risk for pregnancy. This limits the ability of researchers to study those behaviors.

STDs and “outer limits” behaviors

Starting in the late 1980s, these surveys began to focus on another type of risk: STDs, particularly HIV and AIDS. To assess this risk, surveys began to probe respondents for details about previously unmeasured sexual behaviors. Thus, a changing landscape of risk allowed for a particular form of recognition, as surveys began not only to mention but also measure additional forms of sexuality.

In some surveys, this focus on STD risk was made explicit. From 1989 to the present, the GSS has opened its self-administered questionnaire (SAQ) on sexual behaviors with the following statement:

(GSS, 1989 SAQ: 2)

Similarly, right before asking about a variety of sexual behaviors, the NSHS had the interviewer read this statement aloud:

(NSHS, 1996: 75)

Other surveys were less explicit and, instead, simply asked about a “risky” behavior and then immediately asked whether the respondent used a condom (e.g. Add Health, 1996: 400) or had been tested for HIV or AIDS (e.g. NHSLS, 1992: 633). Regardless of whether the motivation for asking was made clear to the respondent, interest in STD risk facilitated questions about previously unmeasured outer limits sexual behaviors. The most common of these, in order of frequency, were same-sex sex, anal sex, and oral sex.

Same-sex sexuality was mentioned in seven of the nine surveys examined (the NLSY and PSID are exceptions), but was not directly measured in most survey-years. Mentions first appeared in the form of opinion questions in 1974, followed by sexual knowledge questions in 1988, both initiated by the GSS. Same-sex sexual behavior questions were introduced in 1989 in both the GSS and NHIS. When surveys were focused on the risk of pregnancy, references to same-sex sex often appeared in the form of lists of behaviors that should not be counted. This instruction from the 1995 NSFG is typical:

(NSFG, 1995: 7583)

Explicit instructions to ignore non-procreative outer limits behaviors are an interesting moment where the surveys both acknowledge that such behaviors exist and fail to count them.

Measurement of same-sex sexual behaviors almost always occurred in the context of STD risk. Although same-sex sex was the most frequently asked about non-procreative behavior, it was measured much less frequently than heterosexual sex precisely because it does not pose a risk for pregnancy. When asking about same-sex sex, these surveys tended to ask respondents about sexual behaviors with both female and male partners.³ For example, the GSS has asked this question every year since 1989:

(GSS, 1989: 3)

This both recognizes bisexual behaviors and allows for researchers using these data to study them. However, this inclusion is likely motivated, in part, by beliefs that male-male sex is particularly risky and that people who engage in bisexual behaviors bridge the gap between the outer limits (gay people) and the charmed circle (straight people), endangering those in the charmed circle (Miller, 1995).

The next most common outer limits behaviors that appeared in these surveys are anal and oral sex. Both first appeared in the NHSLS in 1992, and were mentioned at least once in five of the nine surveys studied. There were more direct questions about anal sex than oral, presumably because anal sex is associated with higher STD risk. However, these behaviors were often not measured directly; instead they were mentioned in definitions of what to count as “sex,” e.g. “by sex we mean vaginal, oral, or anal sex” (GSS SAQ, 1996: 18). These instructions allow for a broader definition of sex, but the resulting aggregate counts make it impossible for researchers to use these data to study oral or anal sex specifically.

Non-consensual sex

In addition to pregnancy and STD risk, surveys also focused on respondents’ experiences of sexual assault. Sexual assault was mentioned in seven of the nine surveys we examined (the NHIS and NLSY are the exceptions), but was directly measured in just 5% of the survey-years (10 of the 190). Questions about non-consensual sex were less frequent than those related to STD risk, but much more frequent than questions about positive aspects of sexuality such as pleasure. About half of the questions measuring non-consensual sex were gendered, portraying males as perpetrators and females as victims, such as this one from Add Health:

(Add Health, 1994–1995: 402)

Questions about sexual assault became more gender neutral over time, such as this one from the 2007 PSID:

(PSID, 2007: 170)

Notably, this single question in 2007 was the only time the PSID asked respondents about sexuality beyond recording marital status. Questions about non-consensual sexual encounters are advantageous, as they allow researchers to study this important topic. However, because these questions are often gendered and are not balanced with questions about sexual pleasure, they reify the belief that sex is dangerous and that men are perpetrators, and women victims, of sexual violence.

Shaping sexual beliefs

Of course, as Vance (1984) argued, sexuality does have dangerous aspects, including unplanned pregnancy, STDs, and sexual violence and we are not suggesting that surveys should not attend to them. However, there are serious consequences to only focusing on the dangerous side of sex. By not balancing questions about danger and risk with questions about pleasure, these surveys reinforce sex negativity. Moreover, by treating certain behaviors as risky, these surveys reify their stigmatized status. Thus, the focus on risk shapes beliefs about sexuality.

The construction of sexuality as dangerous was not limited to the specialty surveys, whose main purpose was to assess health risks. Instead, sex negativity permeated all of the surveys we analyzed. The consequences of sex negativity are exemplified by questions like this one the NSFG asked female respondents:

(NSFG, 1988: 60)

Reflecting a belief that sex is inherently dangerous, this question focused on only one way respondents can make themselves safer: by stopping having sex. It did not ask about other changes, such as starting safer practices like the use of barriers, regular STD testing, or having sex exclusively with other females. Notably, this 1988 question represents the first mention of bisexuals in the NSFG and it treats bisexual men as vectors of disease.

The authorized vocabulary of sexuality in surveys also reproduced beliefs that certain behaviors are taboo. Behaviors that are perceived as risky are almost always asked in an SAQ. This is intended to offer respondents more privacy and anonymity when answering face-to-face surveys. However, by not asking directly about sexual behavior in the main body of the survey, where socially acceptable practices are measured, and by linking particular sexual behaviors with other stigmatized practices that often appear in SAQs, such as drug use and suicide, these surveys further stigmatize non-normative sexual behaviors. Thus, Rubin’s concept is literally mapped onto surveys, with the main survey reflecting the charmed circle and SAQs representing the outer limits.

What surveys miss

The focus on risk in our sample of surveys not only fails to reflect all facets of the sexuality prism, it also distorts the facets that are measured. Sexual behaviors not deemed “risky” were disregarded. Attending to risk also encouraged the measurement of frequency—recording the quantity, rather than the quality, of sex. Lastly, when these surveys measured outcomes of sexuality, they emphasized danger rather than pleasure. These omissions result in an incomplete picture of sexuality.

Safer sex

Available survey data on sexuality are limited to aspects seen as putting respondents “at risk” for specific sexual outcomes of interest. When surveys attend to risk for pregnancy, this means a plethora of data on sex between males and females and a paucity of questions about same-sex sex. Similarly, when surveys ask about sexual behaviors in order to assess STD risk, they rarely ask about behaviors that are not perceived to increase risk for STD transmission. With the exception of the NSHS and NHSLS, the surveys we examined almost never measured masturbation, use of sex toys, or manual stimulation of the genitals. This limits knowledge about safer forms of sex, especially given that data from the NSHS and NHSLS are now more than two decades old.

Moreover, when asking about “risky” behaviors, these surveys often focused only on the “riskiest” forms. Although anal sex and same-sex sex are seen as high-risk behaviors and so are among the most frequently asked about of the outer limits behaviors, not all forms of anal and same-sex sex are equally risky. For example, there are a wide variety of ways to have sex involving anuses, including with tongues, fingers, dildos, plugs, and/or vibrators, but these surveys only measured one form:

(NHSLS, 1992: 563)

By only measuring the particular form of anal sex seen as most at risk for STD transmission, these sorts of questions limit research on the diversity of anal sex behaviors. Anal sex in which a female stimulates a male’s anus and anal sex between females both present low STD risks. Therefore, most surveys do not measure these forms.

The focus on risk in these surveys not only limits the types of behaviors measured, it also limits information about the types of people engaging in measured behaviors and the roles they can take. Notably, the instructions for the NHSLS question about anal sex both limited the definition and specified which couplings should be asked about the behavior, explicitly excluding female/female pairings. In general, many sexual behaviors between females were not measured in these surveys because they are seen as low risk. As stated in the NSHS: “By definition, female-female sexual practices are not risky for HIV transmission. Therefore, lesbians and female bisexuals…are excluded from this risk assessment” (NSHS, 1996: 6105).

Focusing on the “riskiest” factors of sexuality also shapes available data about STDs. Although STD risk was clearly of interest, not all STDs garnered the same attention. As the HIV/AIDS crisis was the impetus for many surveys to begin asking about outer limits behaviors, questions about HIV and AIDS, including those measuring opinions, knowledge, risk, and prevalence, continue to be asked much more frequently than other STDs. This hierarchy of perceived dangerousness ignores other more common STDs such as chlamydia and gonorrhea.⁴

Quality

A focus on risk also means that these surveys devote what little space they allotted for measuring sexual behavior to quantifying the respondent’s risk exposure either through counts of sexual partners or the frequency of sex. These questions from the NSFG and Add Health, are typical:

(NSFG, 2002: 221)

(Add Health, 2007–2008: 335)

The ubiquity of frequency questions stands in contrast to the lack of questions about quality of sex. Just 2% of survey-years (4 of 190) included any measure of sexual satisfaction. ANES asked respondents if they were generally “satisfied” with their sex life once in 1972 (pg. 465), Add Health asked this once in 2007–2008 (pg. 330), and the NSHS and NHSLS each asked about sexual satisfaction once (pgs. 135 and 565, respectively). By contrast, the GSS, NLSY and Add Health feature repeated questions about whether respondents are “happy” in their marriage/relationship, comprising nearly one-quarter (45) of the survey-years examined. The repeated use of such questions means it is much easier for researchers to study trends in marital/relationship happiness than it is to study sexual satisfaction. It is clear that these surveys are interested primarily in the quantity of sex, but not the quality.

Pleasure

Consistent with a lack of attention to quality, mentions and measurements of sexual pleasure were rare in these surveys. This was particularly true for the general surveys. Orgasms, love, enjoyable sensations, and sexual pleasure were never mentioned in the NLSY or the PSID. Both the ANES and the GSS have each asked about sexual pleasure only once in their entire history. In the case of the GSS, that was in a question about which “events strengthened your religious faith”:

(GSS, 1988: 61)

However, sexual pleasure also appeared infrequently in specialty surveys with extensive questions about sexuality. This was especially true of the NHIS, which never mentioned sexual pleasure. The notable exception was the NHSLS, which included a number of questions related to pleasure, including this direct question:

(NHSLS, 1992: 565)

Orgasms are also an important outcome to study as part of the full sexuality prism, even if not everyone finds them pleasurable.⁵ There were many mentions and measurements of orgasms in the NSHS and NHSLS, but acknowledgement of orgasm was otherwise uncommon in these surveys. Orgasm was never mentioned in the NHIS, GSS, NLSY, ANES, or PSID. Add Health mentioned orgasm once and the NSFG includes one repeated question that asks male respondents if a female has “touched your penis until you ejaculated” (e.g. NSFG, 2007: 261). The general lack of attention to orgasm is surprising, given these surveys’ focus on risk; one might expect male orgasms to be of interest because they increase pregnancy and STD risk, but even they often went uncounted.

When these surveys did ask about pleasure or orgasms, they usually did so in the context of danger or dysfunction. Add Health asked about pleasure in a section titled “Motivations to Engage in Risky Behavior.” Similarly, Add Health, NSFG, and NSHS asked opinion questions about whether birth control or condoms reduce “physical pleasure” (NSFG, 2002: 2155) or “make it hard…to have an orgasm” (NSHS, 1996: 139). These questions pit pleasure against the protection of birth control and condoms, reinforcing the false belief that safer forms of sex are antithetical to pleasure.⁶ Mentions of pleasure and orgasms also appeared in questions about dysfunction, as when respondents were asked about difficulties they had in either reaching orgasm (e.g. NSHS, 1996: 155) or finding sex pleasurable (e.g. NHSLS, 1992: 617). Thus, even when asking about pleasure, these surveys often focused on negative aspects of sex. The lack of attention to pleasure means that researchers using these surveys will struggle to analyze the positive aspects of sexuality.

Dangerous data

There are several possible explanations for why large social surveys in the USA continue to focus on sexual risk and danger instead of desire and pleasure. These include mutually reinforcing funding priorities on disease control and a long-standing culture of sex negativity. However, building research questions primarily around risk and danger not only shapes what can be known about sexuality from surveys, it also has unintended consequences for what respondents learn about sexuality when taking surveys. Rather than a balanced view of a range of possible practices, survey respondents see a narrow vision of normative options juxtaposed against a set of risky “deviant” behaviors.

It is not surprising to find that the authorized vocabulary of sexuality in these surveys has been one of risk and danger, given that hegemonic American culture is “sex negative,” treating sex as bad and dangerous (Rubin, 1984). Studying the negative aspects of sexuality has a long history in the USA, which leads to a form of “data inertia” (Merry, 2016), as surveys tend to repeat questions to maintain consistency and facilitate the analysis of trends. In our sample, questions and response options in the late 2000s and early 2010s sometimes reproduced understandings of sexuality held in the 1980s, or even the 1950s, when the questions first appeared. Thus, U.S. surveys either continue to accurately reflect the sex-negative hegemonic culture or are ill equipped to measure its decline.

Similarly, the focus on risk echoes dominant paradigms in both public health and population studies. Epidemiological and demographic epistemology is based on quantifying and analyzing risk—not only in the statistical sense of who is included in analyses but also substantively (Riley and McCarthy, 2003). Federal funding continues to prioritize targeting “populations at risk” (e.g. NIMHD, 2016), an official designation used to evaluate research worthiness. Although the intent is to address inequities, it comes at the cost of stigmatizing often already marginalized people—a move that has been criticized both empirically and theoretically (Cruz, 2017; RWJF, 2010; Young and Meyer, 2005). Research attention thus focuses narrowly on controlling the spread of disease rather than supporting research on pleasure or desire, despite evidence that attention to the joys of sex can reduce the dangers (e.g. Higgins and Hirsch, 2007).

In addition to having deep-rooted causes, the patterns we have highlighted here have important consequences. Survey data serve a normative function, limiting visions of the social world to what gets counted (Cohen, 1999). If available data focus on danger, then sexuality research using such data will also focus on danger (for evidence of this, see Jones, 2019). Consequently, policy recommendations related to sexuality will then emphasize protecting the populace from the dangers of sex rather than on increasing pleasure. This includes promoting the reduction or absence of sexual behaviors, rather than encouraging safer behaviors (see, e.g., Keller and Lindberg, 2020, on abstinence-based sex education). Pairing “sexuality” and “health” has allowed for increased production of knowledge on these topics (Epstein and Mamo, 2017), but thus far it has done little to transform existing paradigms of abstinence and risk (Schalet, 2011) toward alternatives that are sex-positive (Harden, 2014), rights-based (Berglas et al., 2014), and/or focused broadly on well-being (Klugman, 2014).

The focus on risk and danger in survey research on sexuality can be seen as an act of productive power (Foucault, 1978). Surveys are typically seen as gathering information from respondents, but they also teach respondents through the norms represented in questions and answer options. Studies have shown that taking a survey can change respondents’ cultural beliefs and practices (Lee et al., 2014; Williams et al., 2006). If survey questions are shaped by sex negativity, beliefs about sex negativity will be reified. Participants may come away from a survey having learned that sex is dangerous, even if they did not believe that before. Moreover, just as a focus on risk limits what surveys ask, it can shape respondents’ understandings of what forms of sexuality are socially acceptable or even possible. For example, narrow definitions of what counts as anal sex may influence what forms of sex are conceivable to respondents. Surveys that never mention tongues, fingers, toys, sex between females, or anal sex where a female partner stimulates a male’s anus fail to indicate that those sexual behaviors are possible. This further reinforces the status of those behaviors as part of the outer limits of sexuality, or simply erases them entirely. Notably, such silence reduces exposure to safer sexual behaviors, the opposite of the purported goals of many of these surveys.

Social science has long shied away from studying joy and pleasure, focusing instead on how the world is a negative, dangerous place. That said, in advocating for improvements to surveys’ treatment of sexuality, we are not proposing ignoring danger. Instead, like Vance (1984), we argue for balance. There are dangers to sexuality, but there are pleasures, too, and research should attend to sexual pleasure with the same dedication it has shown to danger. Such a shift not only would improve survey data, it would also expand what policy responses can be imagined based on such data and increase what respondents can learn about sexuality when they participate in survey research. Studying pleasure is not all fun and games. One can use questions about pleasure to explore inequalities in reported pleasure and the role of social norms and structures in producing those inequalities. As Jones (2020) details in her call for a “sociological theory of pleasure,” both access to and experiences of pleasure are shaped by a variety of social factors, including race, gender, class, age, and ability, in complex and potentially counterintuitive ways.

In sum, it is time to stop collecting data only on sexual danger stemming from “risky” behaviors and their negative sexual outcomes. Instead, surveys must attend to the full sexuality prism and make desire, identities, a broader range of behaviors, and positive, pleasurable sexual outcomes count. To do otherwise is itself dangerous.

Footnotes

Acknowledgements

We are grateful to Devon Magliozzi, Christine Noack, Sasha Shen Johfre, Josee Smith, and Anita Varma for research assistance.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the American Sociological Association’s Fund for the Advancement of the Discipline, the Clayman Institute for Gender Research, the Institute for Research in the Social Sciences, and the Center for Scholarly and Creative Excellence.

ORCID iD

Laurel Westbrook

Notes

Laurel Westbrook is Associate Professor of Sociology at Grand Valley State University, USA. Their research focuses on the inner workings of the sex/gender/sexuality system, including the role of knowledge production in both maintaining and altering the system. Dr. Westbrook is the author of Unlivable Lives: Violence and Identity in Transgender Activism (University of California Press, 2021).

Jamie Budnick is a postdoctoral fellow at the University of Michigan's Population Studies Center. Her work focuses on the social construction of sexuality knowledge, especially the demography of sexuality and politics of population measurement. She earned her PhD in sociology from the University of Michigan in 2020.

Aliya Saperstein is an associate professor of sociology and the Benjamin Scott Crocker Professor in human biology at Stanford University. Her research examines how categories of difference are operationalized in survey research and the consequences of those methodological decisions for studies of stratification and health disparities.

References

American National Election Study (ANES). Available at: http://www.electionstudies.org

General Social Survey (GSS). Available at: http://gss.norc.org/Get-Documentation

National Health and Social Life Survey (NHSLS). Available at: https://www.icpsr.umich.edu/icpsrweb/HMCA/studies/6647/publications

National Health Interview Survey (NHIS). Available at: https://www.cdc.gov/nchs/nhis/nhis_questionnaires.htm

National Longitudinal Study of Adolescent to Adult Health (Add Health). Available at: https://addhealth.cpc.unc.edu/documentation/

National Longitudinal Survey of Youth (NLSY). Available at: https://www.nlsinfo.org/content/cohorts/nlsy79/other-documentation

National Sexual Health Survey (NSHS). Available at: https://prevention.ucsf.edu/resources/survey-instruments-and-scales

National Survey of Family Growth (NSFG). Available at: http://www.cdc.gov/nchs/nsfg/nsfg_questionnaires.htm

Panel Study of Income Dynamics (PSID). Available at: https://psidonline.isr.umich.edu/guide/default.aspx

10.

Baumle

(2018) The demography of sexuality: Queering demographic methods. In: Compton

Meadow

Schilt

(eds) Other, Please Specify: Queer Methods in Sociology. Oakland: University of California Press.

11.

Baumle

Compton

Poston

(2009) Same-Sex Partners: The Demography of Sexual Orientation. Albany: State University of New York Press.

12.

Baumle

(ed.) (2013) Introduction: The demography of sexuality. In: International Handbook on the Demography of Sexuality. International Handbooks of Population. Dordrecht: Springer, pp.3–10.

13.

Berglas

Constantine

Ozer

(2014) A rights-based approach to sexuality education: Conceptualization, clarification and challenges. Perspectives on Sexual and Reproductive Health 46(2): 63–72.

14.

Biblarz

Stacey

(2010) Ideal families and social science ideals. Journal of Marriage and Family 72(1): 41–44.

15.

Bowker

Star

(2000) Sorting Things Out: Classification and Its Consequences. Cambridge: The MIT Press.

16.

Bullough

(1994) Science in the Bedroom: A History of Sex Research. New York: Basic Books.

17.

Burke K (2016) Christians under Covers: Evangelicals and Sexual Pleasure on the Internet. Oakland, CA: University of California Press.

18.

CDC (2019) Sexually Transmitted Disease Surveillance 2018. Atlanta: Centers for Disease Control and Prevention. https://www.cdc.gov/std/stats18/STDSurveillance2018-full-report.pdf

19.

Chadwick

Francisco

van Anders

(2019) When orgasms do not equal pleasure: Accounts of “bad” orgasm experiences during consensual sexual encounters. Archives of Sexual Behavior 48(8): 2435–2459.

20.

Cohen

(1999) A Calculating People: The Spread of Numeracy in Early America. New York: Routledge.

21.

Cruz

(2017) The making of a population: Challenges, implications, and consequences of the quantification of social difference. Social Science & Medicine 174: 79–85.

22.

Diamond

(2008) Sexual Fluidity: Understanding Women’s Love and Desire. Cambridge: Harvard University Press.

23.

Emens

(2004) Monogamy’s law: Compulsory monogamy and polyamorous existence. New York University Review of Law & Social Change 29(2): 277–376.

24.

Epstein

Mamo

(2017) The proliferation of sexual health: Diverse social problems and the legitimation of sexuality. Social Science & Medicine 188: 176–190.

25.

Ericksen

Steffen

(1999) Kiss and Tell: Surveying Sex in the Twentieth Century. Cambridge: Harvard University Press.

26.

Espeland

Stevens

(2008) A sociology of quantification. European Journal of Sociology 49: 401–436.

27.

Ford

Corona Vargas

Finotelli

Jr. , et al. (2019) Why pleasure matters: Its global relevance for sexual health, sexual rights and wellbeing. International Journal of Sexual Health 31(3): 217–230.

28.

Foucault

(1978) The History of Sexuality, Vol. 1: An Introduction. London: Allen Lane.

29.

Freud

(1949) Three Essays on the Theory of Sexuality. Albury: Imago Publishing Company.

30.

Gates

(2011) LGBT identity: A demographer’s perspective. Loyola Law Review 45: 693–713.

31.

Harden

(2014) A sex-positive framework for research on adolescent sexuality. Perspectives on Psychological Science 9(5): 455–469.

32.

Higgins

Hirsch

(2007) The pleasure deficit: Revisiting the “sexuality connection” in reproductive health. Perspectives on Sexual and Reproductive Health 39(4): 8.

33.

Igo

(2008) The Averaged American: Surveys, Citizens, and the Making of a Mass Public. Cambridge: Harvard University Press.

34.

Interagency Working Group on Improving Measurement of Sexual Orientation and Gender Identity in Federal Surveys (SOGI-IWG) (2016) Evaluations of sexual orientation and gender identity survey measures. Available at: https://dpcpsi.nih.gov/sites/default/files/Evaluations_of_SOGI_Questions_20160923_508.pdf

35.

Jones

(2019) Sex is not a problem: The erasure of pleasure in sexual science research. Sexualities 22(4): 643–668.

36.

Jones

(2020) Camming: Money, Power, and Pleasure in the Sex Work Industry. New York: New York University Press.

37.

Keller

Lindberg

(2020) Expanding the scope of Sex Education and the Teen Pregnancy Prevention Program: A work in progress. Available at: https://www.guttmacher.org/article/2020/02/expanding-scope-sex-education-and-teen-pregnancy-prevention-program-work-progress (accessed 1 June 2020).

38.

Klugman

(2014) Complexity versus the technical fix, or how to put sexuality back into sexual health. Global Public Health 9(6): 653–660.

39.

Laumann

Gagnon

Michael

, et al. (1994a) The Social Organization of Sexuality: Sexual Practices in the United States. Chicago: University of Chicago Press.

40.

Laumann

Michael

Gagnon

(1994b) A political history of the national sex survey of adults. Family Planning Perspectives 26(1): 34–38.

41.

Law

(2009) Seeing like a survey. Cultural Sociology 3(2): 239–256.

42.

Lee

Wilton

Kwan

VSY

(2014) Essentializing ethnicity: Identification constraint reduces diversity interest. Journal of Experimental Social Psychology 55: 194–200.

43.

Louckx

Vanderstraeten

(2015) Household and State-istics: Cornerstones of Society in Population Censuses (Belgium, 1846–1947). Social Science History 39(2): 201–215.

44.

Merry

(2016) The Seductions of Quantification: Measuring Human Rights, Gender Violence, and Sex Trafficking. Chicago: University of Chicago Press.

45.

Michaels

Lhomond

(2006) Conceptualization and measurement of homosexuality in sex surveys: A critical review. Cadernos De Saúde Pública 22(7): 1365–1374.

46.

Miller

Ryan

(2011) Design, development and testing of the NHIS sexual identity question. Hyattsville: National Center for Health Statistics. Available at: https://www.impactprogram.org/wp-content/uploads/2011/11/Miller-2011-HHS-report-on-measuring-sexual-orientation.pdf

47.

Miller

(1995) A review: They said it couldn’t be done: The National Health and Social Life Survey. The Public Opinion Quarterly 59(3): 404–419.

48.

Miller-Young

(2014) A Taste for Brown Sugar: Black Women in Pornography. Durham: Duke University Press Books.

49.

Mills

(1997) Discourse. New York: Routledge.

50.

Mishel

(2019) Intersections between sexual identity, sexual attraction, and sexual behavior among a nationally representative sample of American men and women. Journal of Official Statistics 35(4): 859–884.

51.

Moore

Stambolis-Ruhstorfer

(2013) LGBT sexuality and families at the start of the twenty-first century. Annual Review of Sociology 39(1): 491–507.

52.

Nash

(2014) The Black Body in Ecstasy: Reading Race, Reading Pornography. Durham: Duke University Press Books.

53.

National Academies of Sciences, Engineering, and Medicine (NASEM) (2020) Understanding the Well-Being of LGBTQI+ Populations. Washington, DC: The National Academies Press.

54.

Neuendorf

(2002) The Content Analysis Guidebook. Thousand Oaks: Sage.

55.

NIMHD (2016) Sexual and gender minorities formally designated as a health disparity population for research purposes. Available at: http://www.nimhd.nih.gov/about/directors-corner/message.html (accessed 1 June 2020).

56.

Norton

(2013) Surveying risk subjects: Public health surveys as instruments of biomedicalization. BioSocieties 8(3): 265–288.

57.

Nowakowski

Sumerau

Mathers

(2016) None of the above: Strategies for inclusive teaching with ‘representative’ data. Teaching Sociology 44(2): 96–105.

58.

Oakley

(1998) Gender, methodology and people’s ways of knowing: Some problems with feminism and the paradigm debate in social science. Sociology 32(4): 707–731.

59.

Orne J (2017) Boystown: Sex and Community in Chicago. Chicago: University of Chicago Press.

60.

Parker

Gagnon

(eds) (1994) Conceiving Sexuality: Approaches to Sex Research in a Postmodern World. New York: Routledge.

61.

Pieper

Bauer

(2005) Polyamory and mono-normativity: Results of an empirical study of non-monogamous patterns of intimacy. Unpublished manuscript. Hamburg: Research Center for Feminist, Gender, and Queer Studies, University of Hamburg.

62.

Powell

Blozendahl

Geist

, et al. (2010) Counted Out: Same-Sex Relations and Americans’ Definitions of Family. New York: Russell Sage Foundation.

63.

Riley

McCarthy

(2003) Demography in the Age of the Postmodern. New York: Cambridge University Press.

64.

Rubin

(1984) Thinking sex: Notes for a radical theory of the politics of sexuality. In: Vance

(ed.) Pleasure and Danger: Exploring Female Sexuality. Boston: Routledge & Kegan Paul.

65.

RWJF (2010) A new way to talk about the social determinants of health. Available at: https://www.rwjf.org/content/dam/farm/reports/reports/2010/rwjf63023

66.

Savin-Williams

Ream

(2007) Prevalence and stability of sexual orientation components during adolescence and young adulthood. Archives of Sexual Behavior 36(3): 385–394.

67.

Savin-Williams

Vrangalova

(2013) Mostly heterosexual as a distinct sexual orientation group: A systematic review of the empirical evidence. Developmental Review 33(1): 58–88.

68.

Schalet

(2011) Beyond abstinence and risk: A new paradigm for adolescent sexual health. Women’s Health Issues 21(3): S5–S7.

69.

Sexual Minority Assessment Research Team (SMART) (2009) Best Practices for Asking Questions about Sexual Orientation on Surveys. Los Angeles: Williams Institute.

70.

Shiode

(2016) A tale of two studies: The adolescent health story. Available at: http://www.goldengooseaward.org/awardees/add-health (accessed 1 June 2020).

71.

Silva

(2019) Straight identity and same-sex desire: Conservatism, homophobia, and straight culture. Social Forces 97(3): 1067–1094.

72.

Stacey

(2004) Marital Suitors Court Social Science Spin-sters: The unwittingly conservative effects of public sociology. Social Problems 51(1): 131–145.

73.

Sumerau

Mathers

Moon

(2020) Foreclosing fluidity at the intersection of gender and sexual normativities. Symbolic Interaction 43(2):205–234.

74.

Sumerau

Mathers

Nowakowski

, et al. (2017) Helping quantitative sociology come out of the closet. Sexualities 20(5–6): 644–656.

75.

Tepper

(2000) Sexuality and disability: The missing discourse of pleasure. Sexuality and Disability 18(4): 283–290.

76.

Titscher

Meyer

Wodak

, et al. (2000) Methods of Text and Discourse Analysis: In Search of Meaning. Thousand Oaks: Sage.

77.

Undurraga

(2010) How quantitative are feminist research methods textbooks? International Journal of Social Research Methodology 13(3): 277–281.

78.

van Anders

(2015) Beyond sexual orientation: Integrating gender/sex and diverse sexualities via sexual configurations theory. Archives of Sexual Behavior 44(5): 1177–1213.

79.

Vance

(ed.) (1984) Pleasure and Danger: Exploring Female Sexuality. Boston: Routledge and Kegan Paul.

80.

Wang

(2017) U.S. census to leave sexual orientation, gender identity questions off new surveys. NPR, 29 March. Available at: https://www.npr.org/sections/thetwo-way/2017/03/29/521921287/u-s-census-to-leave-sexual-orientation-gender-identity-questions-off-new-surveys

81.

Ward J (2015) Not Gay: Sex between Straight White Men. New York: NYU Press.

82.

Westbrook

Saperstein

(2015) New categories are not enough: Rethinking the measurement of sex and gender in social surveys. Gender & Society 29(4): 534–560.

83.

Williams

(2006) Still missing? Comments on the twentieth anniversary of “The Missing Feminist Revolution in Sociology.” Social Problems 53(4): 454–458.

84.

Williams

Block

Fitzsimons

(2006) Simply asking questions about health behaviors increases both healthy and unhealthy behaviors. Social Influence 1(2): 117–127.

85.

World Association for Sexual Health (WASH) (2008) Sexual health for the millennium: A declaration and technical document. Available at: http://www.europeansexology.com/files/WAS_2008.pdf (accessed 1 June 2020).

86.

Young

Meyer

(2005) The trouble with ‘MSM’ and ‘WSW’: Erasure of the sexual-minority person in public health discourse. American Journal of Public Health 95: 1144–1149.