Sexual Satisfaction and Research Policies

Sexual Satisfaction

Definitions of sexual satisfaction have commonly focused on positive affect. For example, some researchers have evaluated the degree to which an individual is satisfied or happy with the sexual aspect of his or her relationship (Sprecher and Cate 2004). Other definitions have focused on the perceived balance between positive and negative dimensions associated with one’s sexual relationship (Lawrance and Byers 1995). Still, other researchers have defined sexual satisfaction in terms of individual expectations within the sexual domain, including “the degree to which a person’s sexual activity meets his or her expectations” (DeLamater 1991, p. 62). The common theme that unifies these definitions is the assumption that feeling satisfied is a subjective experience and that the appraisal of satisfaction rests within the person and emerges from their idiosyncratic experience.

In survey research, men and women often report being equally sexually satisfied (Henderson-King and Veroff 1994; Purdon and Holdaway 2006). When gender differences have been found, women often report higher satisfaction than men (Colson et al. 2006; Sprecher 2002; for exception, see Carpenter et al. 2009); however, recent critical work in this field of research has argued that women may hold lower expectations for sexual satisfaction and therefore may report higher ratings when assessed in research settings (Carpenter 2010; McClelland 2010). There is far less research on sexual minority sexual satisfaction rates, but existing research with gay and lesbian samples suggests that sex in committed same-sex relationships is similar to heterosexual marital ratings of sexual satisfaction (Deenen et al. 1994; Holmberg and Blair 2009; Kurdek 1991; Lever 1994). Kurdek (1991), for example, found no differences in sexual satisfaction among four types of couples: gay, lesbian, married heterosexual, and cohabitating heterosexuals.

Researchers have asked if sexual satisfaction is defined differently across demographic groups. Early studies found that physical satisfaction was consistently prioritized by men, while women prioritized emotional closeness and intimacy when evaluating their sexual satisfaction (Laumann et al. 1994). Others, however, have argued that gender norms may explain these differences, specifically, gender norms which emphasize women’s emotional capacities and often downplay their investment in physical pleasure (Bliss and Horne 2005; Frith and Kitzinger 2001; Gagnon and Simon 1970; Sanchez et al. 2005). This question of how sexual and gender norms shape the expectations of those groups and individuals who experience limited sexual rights (i.e., those individuals for whom laws, violence, coercion, norms, or biases regulate aspects of their sexual behaviors, relationships, sexual health, and gender presentation; see International Planned Parenthood Federation 2008; Richardson 2000) has become an increasingly pressing question for researchers who study the quality of sexual experiences and relationships across diverse populations (Bliss and Horne 2005; Diamond and Lucas 2004; Fahs and Swank 2010; Impett and Tolman 2006).

The vast majority of sexual satisfaction research has been done using self-report survey designs. There are surprisingly few studies of sexual satisfaction using interview methods. The few that have been done focus separately on either women (Bay-Cheng et al. 2009; Fahs 2011; Nicolson and Burr 2003; Tolman 2002) or men (Epstein et al. 2009). The few studies that have looked at both men and women have found gender differences. For example, in their interview study with heterosexual men and women with varying levels of illness, Daker-White and Donovan (2002) found that men defined their sexual satisfaction in terms of intercourse frequency and the match between this frequency and their libido, while women defined satisfaction in terms of intercourse frequency, trust, and mutual enjoyment.

Looking more closely at the interview-based research with female samples, there have been studies which focused on the psychological processes described by women when they were asked to describe their sexual satisfaction. For example, Nicolson and Burr (2003) found that female participants’ commonly described sublimating their own needs in an effort to ensure their male partner’s sexual satisfaction. This theme has been found consistently in qualitative studies, particularly in research with young women (Holland et al. 2004; Thompson 1995; Tolman 2002). This set of findings has been important in highlighting a set of psychological maneuvers that women describe when evaluating their own sexual desire or satisfaction. What remains underexplored is the way these translate into research using survey items, and importantly, the degree to which this consistent finding has been understood by researchers evaluating sexual dysfunction—a field which often relies on sexual satisfaction ratings as important indicators of well-being (e.g., Arrington et al. 2004). Additionally, the bulk of this qualitative research has involved women reporting on sexual encounters with male partners—much less is known about sexual relationships involving same-sex partners, for both men and women.

Subjectivity and Satisfaction

In their overview of research on sexual and relationship satisfaction, Schwartz and Young (2009) discussed how research in the field of satisfaction often presumes that, “everyone knows what it means to say that someone is satisfied or dissatisfied” and few researchers actually define the term in studies (p. 1). Schwartz and Young argue that this lack of an agreed-upon definition, combined with the diversity of satisfaction definitions, has encouraged researchers to rely on subjective assessments rather than the more specific behavior- and emotion-based items, such as “Do you wake up in the morning ready and willing to face the day?” as seen in depression and quality of life studies. Schwartz and Young argue that these types of items which isolate specific elements from the psychological construct are more or less unknown in the sexual satisfaction literature.

While this is the case and may continue to be the trend in sexual satisfaction research, Schwartz and Young’s comparison to these other bodies of research alerts us to something important: much like research on depression and quality of life, which rely on assessments to guide clinical decisions and interventions, sexual satisfaction ratings are now similarly used in clinical settings. While sexual satisfaction was (and still is) considered an interesting and potentially useful way to predict, for example, relationship continuity (e.g., Sprecher 2002), it has increasingly become a highly valued clinical outcome that is used in medical settings and increasingly interpreted in terms of cut scores and clinical thresholds (Dennerstein et al. 2006; Tunuguntla 2006). This shift in the importance of this psychological variable requires that researchers not merely rely on the idea that satisfaction is subjective by nature, that everyone knows what it means, or that there is some agreed-upon definition that guides satisfaction appraisals (see McClelland 2010).

Clinical Endpoint of Interest

While sexual satisfaction seems an unlikely candidate for influencing any type of public policy, this psychological construct has become of great interest to researchers and policy makers working in the field of female sexual dysfunction over the last decade. In 2000, the US Food and Drug Administration (FDA) released their “Guidance for Industry,” in which they recommended that investigators conducting clinical trials of pharmaceuticals used to treat female sexual dysfunction use the number of “successful and satisfactory sexual events [SSEs] or encounters over time” as the “primary endpoints” (i.e., dependent variables used to assess clinical effectiveness) in their studies (FDA 2000, p. 3). In other words, in order for a sexual dysfunction drug to be proven effective, investigators must demonstrate that participants’ reported sexual satisfaction increased over time, or more specifically, that they report an increasing number of satisfying events. Sexual events are defined as including “satisfactory sexual intercourse, sexual intercourse resulting in orgasm, oral sex resulting in orgasm, and partner-initiated or self-masturbation resulting in orgasm” (FDA, 2000, p. 3). Two important points stand out in this series of descriptors: the importance attributed to an individual’s satisfaction rating and the conflation of satisfactory and orgasmic sex.

The FDA recommendations made a second point in their guidance worth noting. It states, “[t]he determination of successful and satisfactory should be made by the woman participating in the trial, as opposed to her partner” (FDA, 2000, p. 3). While this statement marks the FDA’s attempt to ensure that determinations of sexual satisfaction are consistent across studies, and perhaps an attempt to guard against women deferring to (male) partners’ sexual evaluations, its presence in the FDA guidance signals to the reader (and researchers) that gender norms are acknowledged, but without the why or how these norms operate. The language implies an invisible and unnamed other who may usurp the woman’s decision, as well as a female figure who is (potentially) unable to speak for herself. Also left silent is how investigators should guard against this happening in their studies.

In a letter to the FDA that same year, as part of the New View Campaign—a group of clinicians and social scientists advocating for a classification system of female sexual dysfunction that in addition to physiological pathology, also considers social, political, and interpersonal factors in a woman’s life—Leonore Tiefer and 21 colleagues outlined their concerns about the use of sexual satisfaction as a primary endpoint as it had been outlined in the FDA (2000) clinical trial guidance (see Tiefer 2001). In the letter, Tiefer and the letter’s co-signers argued that orgasm was a poor indicator of sexual satisfaction and argued that, “Women frequently evaluate the desire for and the pleasure and intimacy of physical experience in relation to emotional issues such as safety and satisfying their partner” (Tiefer 2001, p. 74). The letter went on to argue that the FDA’s directive about the determination of satisfactory coming from the woman herself was unrealistic given gender norms, female socialization, and heterosexual relationship dynamics (Tiefer 2001). What the letter was not able to include was empirical evidence examining the processes men and women use to make sexual satisfaction appraisals in research settings.

Flibanserin Example: A Case Study

In 2010, Boehringer Ingelheim petitioned the FDA for approval of their drug flibanserin, a pharmaceutical drug that was being tested as a treatment for female sexual dysfunction (Boehringer Ingelheim 2010). In alignment with FDA guidelines, one of the primary end points used by Boehringer Ingelheim’s staff was the count of “satisfying sexual events” (SSEs) that women noted in randomized, double-blind, placebo-controlled trials. In their 2010 briefing report, Boehringer Ingelheim measured this outcome as, “the change from baseline to the final visit period in the monthly frequency of SSEs as measured by the eDiary question: ‘Was the sex satisfying for you?’” (p. 31). Responses were measured dichotomously as “yes” or “no.”

The pooled analysis detailed in the report released by Boehringer Ingelheim included 971 (flibanserin 100 mg, n = 450; placebo, n = 521) pre-menopausal women who completed the 24-week trials. In that analysis, a daily 100-mg dose of flibanserin significantly increased the frequency of sexually satisfying events versus the placebo. The size of that increase, however, was quite small. Women who received the drug reported an increase of 2.1 satisfying events over the 24-week study period, while those who received the placebo reported an increase of 0.9 events. This translates to, on average, a little more than one (1.2) satisfying events per month for women in the medication group over and above what they would have reported on placebo. How women defined what was satisfying was left up to them. This follows the trend in satisfaction research more broadly where the determinants for satisfaction are left to the participants and not externally imposed (Diener et al. 1999, 2003). However, some have argued that reliance on subjective appraisals in domains where individuals experience unequal rights, such as the sexual domain, obscure the fact for some individuals and groups, sexuality is defined as dirty, immoral, dangerous, and even illegal (Diamond and Lucas 2004; McClelland 2010). As a result, satisfaction appraisals made from within these sociopolitical contexts may not be comparable with those ratings made by peers who do not face these same potential legal and social costs associated with being sexual.

The Boehringer Ingelheim staff addressed the issue of subjective measurement by anchoring each woman’s response to her own baseline measurement and analyzing change over time for each participant. This method, they explained, would address the issue of women having different thresholds for satisfying sexual events. They wrote:

This anchor-based assessment of response is an important clinically relevant endpoint, because a result on any given measure that is statistically significant may or may not represent what is truly meaningful to an individual woman taking flibanserin (Boehringer Ingelheim 2010, p. 20).

Even with these methodological attempts to create reliable and valid measures of sexual satisfaction on the part of flibanserin’s manufacturers, questions remain about several aspects of this research. For example, we know little about the point at which each woman anchored her baseline measurement: What criteria were used to anchor the original assessment? We know little about the overall meaning of “satisfying” when applied to sexual events for this group of women: What were the outcomes, feelings, and experiences that were linked with an individual determining that an encounter was in fact, satisfying? Lastly, we know little about role of partners in these determinations of satisfaction: What was the role of another (either present or imagined) when determining whether an event was sexually satisfying?

Sexual Satisfaction: An Empirical Analysis

Taking both the 2000 FDA guidance and the 2010 example of flibanserin together, three important points are raised that are relevant to researchers and policy makers working in the fields of sexual dysfunction, as well as sexuality research more broadly: (1) the definition of sexually satisfying is not made explicit, yet is relied upon as the main outcome of interest when determining the effectiveness of clinical interventions; (2) satisfaction is, at times, conflated with orgasm, but the conceptual and experiential overlap between the two is unclear; (3) satisfaction appraisals are described (and mandated) as coming from the woman and not her partner, but little is known about the psychological process involved in making a sexual satisfaction appraisal. The first two issues concern definitions (what does it mean to say one feels satisfied?) and the third issue concerns the genesis of the appraisal (whose experience is considered during a satisfaction appraisal?). These three issues were examined in a study of diverse young adults. Participants were asked to both evaluate their sexual satisfaction in a series of survey items and narrate their sexual satisfaction appraisals during a semi-structured interview. This multi-method design enabled an analysis of several research questions: (1) whose experience is salient when individuals report on their own sexual satisfaction; (2) what do survey item and interview data enable us to see on their own and in unison; (3) do men and women of varying sexual identities use similar criteria to evaluate their level of sexual satisfaction?

Methods

Sample and Recruitment

Participants were recruited from the Psychology Research Pool at a diverse undergraduate college in New York City in 2008. Potential participants responded to an on-line ad for a study concerning “dating and relationships.” The ad specifically did not mention sexual satisfaction in order to reduce potential sampling bias in the case that only those who were highly satisfied or dissatisfied would decide to participate. In addition, the call for participants explicitly named LGBT and heterosexual relationships and stated that participants did not need to be in a current relationship to be eligible to participate. This decision was made in order to not limit the sample to individuals who were engaged in sexual relationships; masturbation was considered a relevant form of sexual expression in this study.

Eligibility criteria were over 18 years old and able to speak and write English fluently. All potential participants were asked to report several demographic details during the screening process and were screened for those that responded “yes” to an item that asked them if they were “sexually active.” In all, 386 people were screened, 375 were eligible, and 79 individuals were invited to participate in the study. Of those who responded to email and phone requests to schedule a face-to-face appointment, recruitment continued until there were six to ten participants for each of the four groups of interest (heterosexual men, heterosexual women; sexual minority men and sexual minority women), a sampling procedure recommended for qualitative studies involving heterogeneous samples (Guest et al. 2006; Morse 1994). The final sample consisted of 41 individuals (51% women, 44% LGBT; 54% white; mean age of 23.2; see Table 1 for demographic characteristics of the sample).

Table 1 Demographic characteristics of sample
Full size table

Study Procedures

Once participants consented to be a part of the study, they completed a semi-structured interview and survey items. Each participant was interviewed by a female researcher (the author). Interviews typically lasted about 25–35 min, were audio-recorded and transcribed for analysis. In any mixed methods study, research design must account for the sequence in which the methods are administered. While there are many arguments made by researchers using mixed methods designs as to the sequence in which qualitative and quantitative methods are implemented (Creswell et al. 2003; Ivankova et al. 2006), there is less research on the psychological effects of sequencing within a cross-sectional design. The concern in this study was less about which method was prioritized by the researcher (see Creswell et al. 2003), but rather, on the experience of the participant and how sequencing might affect his or her language choices, priming and saliency of specific ideas, and the level to which he or she felt comfortable expressing experiences outside the boundaries implied by the survey item. As a result, the sequence was consistently semi-structured interview, followed by self-report survey items. Starting with the interview encouraged participants to think globally about the construct of sexual satisfaction and gave them permission to hold contradictory and ambivalent experiences, attitudes, and opinions about sexual satisfaction. The reasoning was that if the survey items had been administered first, participants may have felt obliged to maintain the non-ambivalent attitudes that survey items often require.

For the purposes of this study, sex was defined broadly and included masturbation, fondling, caressing, intercourse, oral/genital contact, and genital contact with another person(s). This definition was crafted in order to ensure that intercourse was not assumed to be the only form of sexual expression relevant in the study. This decision was made so that all participants, regardless of sexual orientation, would consider sexual activities other than penile–vaginal intercourse as potentially relevant to the study. Participants were reminded of this broad definition both in writing and out loud at the beginning of the interview and survey portion of the study.

Semi-Structured Interview

Semi-structured interviews were conducted in order to collect participants’ experiential data and to enable participants to feel comfortable describing their own experiences with intimate relationships and sexual activities. The interview protocol contained a number of specific questions, but rather than follow a strict set of probes, the interviewer responded to each participant’s stories and explanations with questions designed to elicit each person’s idiosyncratic definitions and experiences (Conrad and Schober 2008). The interview questions did not ask participants to elaborate specific sexual experiences, but instead, to describe the way that they interpreted these experiences (past and present) and to describe how their experiences helped them distinguish satisfactory from unsatisfactory sexual experiences. The interview protocol included questions that asked participants to describe whether they thought about their own sexual satisfaction, their definitions of sexual satisfaction, criteria they used to decide if they were sexually satisfied, previous or current sexual experiences that influenced their sexual development, what they expected in terms of sexual experiences, as well as any developmental changes they observed in themselves in the recent past and any changes they anticipated in the future. If they thought about their own satisfaction, they were asked to discuss how long have they have thought about it, whether these definitions have changed over time and if so, when, and were there were circumstances that prompted these changes.

Survey Items

Participants were asked three survey items using a pencil and paper format. Each was asked to report on his or her overall sexual satisfaction using a modified version of Cantril’s Ladder (Cantril 1965). Using an un-anchored 10-point scale, participants were asked, “How would you rate your overall level of sexual satisfaction? Please mark your response anywhere on the line below” (n = 40). Frequency of orgasm was measured using a single item, “When you and your partner have sexual relations, how often do you have an orgasm—that is, climax or come?” Responses ranged from 1 (never/hardly ever) to 5 (most of the time/every time). The extent to which participants reported liking sex was measured using a single item, “How much do/did you like having sex (of any type) with your partner?” Responses ranged from 1 (dislike very much) to 5 (like very much). These last two items came from the National Longitudinal Study of Adolescent Health (Add Health; Harris et al. 2009) and were part of a larger study; as a result, only participants who indicated that they had a partner were asked to report on their frequency of orgasm and the extent to which they liked sex with their partner (n = 28). Items were standardized prior to analysis.

Analysis

Narrative Analysis

Given that the research questions that drove this study concerned the subtle ways that individuals might encounter and narrate experiences with intimate others, as well as usually unarticulated dimensions of their psychological experience, a narrative analysis approach was used. Narrative approaches have unique strengths when examining individual processes in the midst of social norms (Billig et al. 1989; Gergen 2001; Lieblich and Josselson 1994). The analysis focused on the ways that participants imagined their sexual expectations and the quality of their sexual experiences by focusing on the benchmarks, decisions, and criteria that individuals used to determine the degree of their sexual satisfaction. Narrative, rather than thematic or content analysis strategies, were preferable in this case because the point of view of the participant was considered primary, rather than, for example, reported sexual behaviors or outcomes. The focus in the current study was on the subjective experience of individuals, how they related their evaluation process, and how social contexts (particularly those relating to gender and heteronormative expectations) shaped participants’ narratives.

Coding Procedures

The 41 semi-structured interviews were coded by the author using NVivo 9 to organize the transcript data, as well as coding throughout the analysis phase. Coding procedures focused on two aspects of participants’ narratives: relationality and benchmarks.

Relationality

Narrative researchers (Burkitt 2010; Freeman 2007; Raggatt 2007) have argued for greater attention to relational narratives in the study of lives. This emphasis on the relational shifts the focus from theorizing a singular self who recounts his or her own life and experiences, to a more complex web of theories and methods that can attend to the multiple others that influence a narrative. For the purposes of this study, relationality was defined as “an emphasis on dynamic processes of connections and transactions, as opposed to substances and isolated individuals” (Pachucki and Breiger 2010, p. 207) and concerned how participants included others within stories of the self (Bakhtin 1981; Hermans 2001, 2002). Researchers (e.g., Pachucki and Breiger 2010) have argued that this type of analysis is essential to understand how individuals’ multiple meanings and expectations change over time and in response to social cues from both proximal and distal others, ranging from intimate partners to gendered sexual norms and sexual stigma (Herek 2007).

Benchmarks

Those parts of the interview transcript where individuals narrated how they evaluated their level of sexual satisfaction were coded for the content of the benchmarks (i.e., did a participant name a specific feeling as primary satisfaction criterion?), as well as the process they used to narrate these criteria during the interview (i.e., were criteria easily imagined or more difficult for the participant to imagine?). This included any and all references participants made to how they decided whether they were sexually satisfied and to what degree. Descriptions often included physical, cognitive, affective, or relational evidence used throughout the interview to describe how participants imagined and processed their own sexual satisfaction appraisal. In addition, the sequencing of this evaluation was considered (i.e., did they first consider their own experience and then their partner’s?).

Data Triangulation

Pairing survey data with interview data has proven effective in examining the strengths and weaknesses of available measures (Galasiński and Kozłowska 2010; Sudman et al. 1996). Given the context and growing importance of sexual satisfaction evaluations, observing how the two datasets converge and diverge provided useful insights into the basic question concerning the level of agreement between the two data types. This insight helped to evaluate whether self-report items sufficiently capture psychological elements of intimate relationships, sexual expectations, and meanings attached to “feeling satisfied” more broadly.

This comparison was accomplished using a triangulation design, defined as a study that collects and analyzes the quantitative and qualitative data during the same phase of the study, as opposed to using these data sequentially (Plano Clark et al. 2008). Triangulation designs in mixed methods studies are used when the objective is to compare and contrast the results from each data source or by synthesizing what is learned from each (e.g., Currie 1998). In the current study, both data collection methods were designed to address the same concept—sexual satisfaction—and it is in their comparison that we are able to observe the types of information that are communicated in each method.

Rather than transform the interview data into a quantitative group-level variable (which may have inaccurately represented the qualitative data as static), when an individual’s interview is excerpted, their excerpt was paired with their response to the survey item concerning their overall sexual satisfaction. When interviews are excerpted below, participants’ demographics are provided, including self-indentified gender, sexual minority status, and age, as well as how they rated their overall sexual satisfaction on the previously described un-anchored 10-point scale (this number is noted by “rating” following each participant’s quote).

Findings

Survey Data

The survey items in this study enabled analysis of potential group differences: did men and women—of varying sexual identities—report similar levels of sexual satisfaction? The unstandardized means and standard deviations are presented in Table 2. A series of two-way analysis of variance tests assessed the effect of gender and sexual minority status on orgasm frequency, liking sex with a partner, and sexual satisfaction. The analysis for orgasm frequency showed no significant main effect for gender, F(1,24) = 0.003, p = 0.96; no significant main effect for sexual minority status, F(1,24) = 0.003, p = 0.96; and no significant interaction between gender and sexual minority status, F(1,24) = 1.88, p = 0.43. This same pattern was observed for liking sex with a partner: no significant main effect for gender, F(1,24) = 0.009, p = 0.93; no significant main effect for sexual minority status, F(1,24) = 1.50, p = 0.23; and no significant interaction between gender and sexual minority status, F(1,24) = 0.79, p = 0.38. This same pattern was also found for overall sexual satisfaction: no significant main effect for gender, F(1,36) = 0.05, p = 0.83; no significant main effect for sexual minority status, F(1,36) = 1.30, p = 0.26; and no significant interaction between gender and sexual minority status, F(1,36) = 0.80, p = 0.38. These findings suggest that these four groups reported approximately the same rate of orgasm frequency, liking sex with their partner, and roughly the same level of sexual satisfaction.

Table 2 Means and standard deviations for orgasm frequency, liking sex, and sexual satisfaction
Full size table

Researchers have often measured the frequency of orgasm as one of the dimensions of sexual satisfaction (Barrientos and Páez 2006; Fugl-Meyer et al. 2006). The findings from the current study demonstrated that these two constructs had a moderate association. For participants who answered both items, orgasm frequency and sexual satisfaction were significantly correlated, r(26) = 0.42 (p < 0.05). Sexual satisfaction was also significantly associated with liking sex with one’s partner, r(26) = 0.40 (p < 0.05). Orgasm frequency was also significantly correlated with liking sex with one’s partner r(26) = 0.51 (p < 0.01). As these samples sizes are small—only those currently partnered answered orgasm and liking sex items—these correlation data should be interpreted as exploratory. From this small sample, it does appear that that these three constructs overlap, but are not identical.

Interview Data

The main research question driving the current study concerned the ways that individuals appraised their sexual satisfaction, with particular focus on the subjectivity of the speaker during the appraisal process. The narrative analysis revealed important aspects of the participants’ varying use of their own subjectivity in their appraisals: While most participants relied on their own experience as the main source of information when evaluating their level of satisfaction, some participants stepped away from their subjective experience and, instead, relied on the experience and evaluation of their partner when making a satisfaction appraisal.

With this in mind, the findings from the interview portion of the study are presented in two groups: participants whose descriptions relied on their own experience as the primary benchmark (the assumed process in self-report research) and participants who relied on a partner’s experience when determining sexual satisfaction. These groups should not be interpreted as indicators of participants’ stable identities. These data, however, demonstrate the variety of ways that participants structured their responses to questions about sexual satisfaction.

Satisfaction Anchored in the Self

All of the men in the study reported that having an orgasm was an important benchmark for determining sexual satisfaction. A typical example of this sentiment is in the following statement when a male participant was asked how he judged whether he was sexually satisfied: “I mean basically, having an orgasm basically” [male, heterosexual, 23, rating 8.00]. However, there were important variations within the male participants, particularly in terms of how much the participant weighted the importance of both partners having an orgasm in a single sexual encounter. For example, the following young man explained that the shared quality of both him and his partner having an orgasm was important to him when evaluating his satisfaction level: “It was very satisfying, and because I, I felt like it was a good balance, it was satisfying for both of us, it was, like, we both were able to have orgasms, like that was, it was equal, and that, that felt good to me” [male, gay, 22, rating 8.00].

In contrast to the men in the sample, no women described their own orgasm as the primary benchmark for their own satisfaction. They did, however, use a number of other criteria which were described as being anchored in their own experience, including feeling close to one’s partner. As this woman explained: “Like it’s satisfying even if you know like that realistically you’re not having an orgasm it’s still really satisfying just to kind of be with that person, you know?” [female, heterosexual, 29, rating 6.5]. In this description, an orgasm is imagined as potentially preferable, but not a necessary component for her to feel satisfied. However, this speaker also indicates that a non-compensatory processFootnote 1 may be at work. She relates in this short statement, a series of decisions that are put into play in the midst of a sexual experience: if orgasm is not possible, then “just to kind of be with that person” becomes her new benchmark for feeling satisfied. This example demonstrates that individuals are constantly making decisions about what they are going to consider “good enough” within their sexual experiences and that a sexual satisfaction appraisal is the product of a complex set of prioritizing both possible and probable sexual outcomes. Orgasm, for her, is described as an ideal, but perhaps an improbable threshold, so just being with her partner becomes the new threshold she uses to determine her satisfaction level.

In addition to feeling close to a partner, there were some participants for which achieving a sexual encounter was considered satisfying. In the quote below, a young woman described the overall experience of sex to be the benchmark for her satisfaction. For this participant, enjoyment and satisfaction were embedded in the process of sex itself, “whatever it might be.” In this case, the act of being sexual itself was found to be the necessary criteria. She explained:

I find the act of sex to be satisfying. So it’s not necessarily the end point of the orgasm that should be satisfying, that to me is kind of irrelevant, it’s nice, and it’s a part of it, but I just enjoy sex in general. Whatever it may be, I just enjoy it. So that is satisfaction for me [18, bisexual, female, rating: 8.5].

In contrast to descriptions which highlighted the physical or emotional outcomes of a sexual encounter, some participants described the level of comfort they felt with a partner as one of the important elements when evaluating their satisfaction. While still a relational aspect of sexuality, these descriptions relied less on the emotional connectedness to a partner and more on the partner’s role in the participant’s life. For example, a young man described his benchmark for satisfaction in terms of the amount of labor it took to meet someone and the greater level of comfort he felt with known partners. In the quote below, he begins by saying that sex is satisfying if he doesn’t have to “do anything” to meet a partner and then goes on to say that sex is more satisfying with partners that he knows as opposed to partners he has not met before. In the quote, we see a relationship between the idea of comfort and satisfaction.

Well, I guess the act of sex is satisfying enough that I don’t have to do anything. I don’t have to go to the restroom somewhere, and meet some random person online. I mean, it doesn’t have to happen that way for me to feel sexually satisfied, but, you know, I don’t know. I guess I feel more comfortable with a person I’ve known for a while. That’s like one of the leading things that I would feel are more sexually satisfying than just some random person [18, gay, male, rating: 6.5].

While we see evidence of different ways that men and women prioritized experiences to form their satisfaction appraisals, the aspect that unites this first group of responses is that they were all appraisals that were anchored within the self. These ranged from physiological responses, such as orgasm, to emotional responses produced by feeling close to or comfortable with a partner. These subjectively derived responses stand in contrast to those that were anchored outside of the self and in the body or experience of the partner.

Satisfaction Anchored Outside of the Self

When the interview data were analyzed for the genesis of the sexual satisfaction appraisal (i.e., where the decision originated), several patterns were evident. Instead of looking inward, a sub-sample of participants described looking to their partners’ experience for indications of how satisfied the participant felt. This ranged from looking for evidence that they had fulfilled their partner’s wishes or desires, to substituting their partner’s appraisal for their own.

For some women who had never experienced an orgasm, their partner’s orgasm provided a benchmark by which to decide whether the sexual encounter was satisfying. For example, this young woman’s description of using her male partner’s orgasm as her point of reference also makes evident a set of non-compensatory decisions when she described her own orgasm as “not that big of a deal”:

Well, right now, like, I’ve never had an orgasm, and it’s not, I guess, just from not knowing what it feels like…I mean, [orgasm] is not that big of a deal. …Well, I just, I like to know that the person that I’m with that, like, they’re satisfied. You know, at least that they’re having an orgasm and that I can satisfy whatever it is that they want [female, heterosexual, 19, rating: 4.5].

This participant equated orgasm with satisfaction and since she did not consider her own orgasm important, she used her partner’s satisfaction as a proxy for her own satisfaction. A similar use of the male partner’s level of satisfaction can be seen in the next example. What differentiates these two responses is that this second woman is not deferring to her partner because she does not experience orgasm, but because she does not define feeling sexually satisfied as important to her:

Um, I don’t really think that much about sex I guess so, yeah, for me I mean getting sexually satisfied isn’t a big priority for me. It’s mainly him I guess. The number one priority is for him to be sexually satisfied. Not for me. I guess it’s easy for me to be sexually satisfied, since like if he’s sexually satisfied then I’m sexually satisfied [female, bisexual, 18, rating: 7.0].

The explicit description, “if he’s satisfied then I’m sexually satisfied” is key—not only in terms of being evidence of deferral of female sexual agency, but also because it is evident that measures of sexual satisfaction may not be accurately capturing this use of a proxy when making sexual satisfaction decisions. This may be evident that while self-report survey items appear to be reflections of the self, they may in fact be reflections of the partner or some combination of participant and partner. This sentiment of wanting to satisfy a partner was expressed by more than half of the female participants. Some explained that their partner’s satisfaction was used as their primary benchmark—ultimately lodging the participants’ satisfaction within another person.

Reliance on a partner’s experience was not limited to women in this study. LGBT and queer-identified men also described fulfilling a partner’s expectations as the primary benchmark used to evaluate their own level of sexual satisfaction. At another point in the interview, the young man quoted below described himself as the “more feminine partner” in the couple and associated this role with specific sexual responsibilities within his same-sex sexual relationship:

I feel that if I can, if I can do what’s expected of me, I feel that I have been rewarded something…I feel like as a partner, as being someone’s boyfriend, as someone’s significant other…I feel like it’s my duty, or that it’s their duty as well, to satisfy one another. But mostly I feel like it’s my job to do so. I want to see that my partner is happy, I want to see that my partner is feeling great, that they’re satisfied [male, gay, 21, rating: 6.5].

While this example explicitly links gender roles to sexual appraisals, other examples from LGBT and queer-identified men described a similar sentiment of sexually satisfying male partners as “work” and as a “duty.” This pairing of feminine responsibilities in the private sphere (and the intimate sphere more specifically) with images of work is provocative evidence of how same-sex couples struggle with gendered expectations both in and out of sexual relationships and how gender roles organize both male and female bodies (Marecek et al. 1983; Peplau and Spalding 2000).

Summary

Differences were observed in how participants made decisions about their own sexual satisfaction. While some used subjectively derived criteria, others looked outside of themselves and, instead, towards their partner in order to narrate their level of satisfaction. The interview data revealed that while the presence of orgasm was used as a benchmark, it was by no means the only benchmark used by participants. The other benchmarks here were highly embedded, not only within relationships, but within the partners themselves. Partners’ satisfaction was sometimes used as a proxy for one’s own satisfaction, how close one felt with a partner was often used as a means to evaluate the quality of the sexual relationship, and “doing a good job” was used as a benchmark for assessing whether the sexual activity was considered successful—an alternative reading of satisfying. In sum, the presence of an orgasm was used mainly by men, both heterosexual and LGBT, as the criteria to decide their satisfaction level. Women, on the other hand, did not rely on their own orgasm as their main criterion, but instead used their partner’s orgasm or relied on less physiological indicators to judge their level of satisfaction. LGBT and queer-identified men also looked to the relational dynamic for evidence that they satisfied their partner.

Pairing Survey and Interview Data

By pairing survey ratings with participants’ interview data, it was possible to observe several patterns about how individuals made satisfaction appraisals. In this study, satisfaction ratings above and below the sample mean were found in both those who anchored their satisfaction appraisal in the self and those who anchored their satisfaction outside of the self. Looking to those individuals who reported high levels of sexual satisfaction, we saw a diversity of criteria used to describe what is considered sexually satisfying. For example, those individuals who evaluated their sexual satisfaction above the sample mean (M = 7.30, SD = 1.61) described orgasm as an essential component of satisfaction, described the act of sex itself to be satisfying, and included those who described the experience of both partners having orgasms as what was considered satisfying for them. Those who reported satisfaction levels below the sample mean ranged from individuals who reported being satisfied by just being with their partner, to feeling more satisfied in encounters with known rather than random partners, and finally, to those individuals who reported that they satisfied their partner in order to feel satisfied themselves.

Discussion

Overall, this study allowed for a closer examination of several important questions that are infrequently considered in psychological research: When participants provide self-report data in research settings, do these data actually reflect the evaluation of the individual? What are the research and policy implications if self-report data are not anchored in the self, but anchored in the experience of another person? Lastly, what assumptions about the “self” are implicit in self-report measures in psychological research more generally? Below, I discuss two patterns observed in the current study which highlight how simply describing these evaluation processes as “subjective” may obscure other social psychological qualities of these appraisals.

Satisfaction by Proxy

When individuals are asked to rate their own sexual satisfaction, researchers assume an intra-individual reflective process in which the person looks inward and decides how satisfied he or she feels. Interview data from the current study revealed that this process is far more inter-individual than previously thought. When individuals report on “my sexual satisfaction,” this response represents more than just a singular “me” for some. This shift in perspective results in responses that are determined by a combination of me, them, and us.

Feminist psychologists have long argued for a more sophisticated analysis of how individual selves formed in social environments are never without the influence of the social—even when alone. Holland et al. (2004) reported similar findings in their interviews with young women in which they found that young women defined their own sexual satisfaction in terms of a “general contentment with the relationship, in which her sexual satisfaction is limited or regarded as unnecessary” (p. 110). They described this phenomenon as “male-in-the-head,” which they defined as that which “regulates the expectations, meanings and practices of both men and women” (p. 156). The current study extends this earlier finding by demonstrating that not only did the participants defer to male partners, but in fact used their male partner’s evaluations in place of their own. This study also demonstrates that this dynamic may not be limited to heterosexual women, but in fact may be applicable to same-sex male couples as well. There was evidence that some sexual minority male participants looked to their male partner’s level of satisfaction instead of their own. This finding may speak to other research which has examined the influence of gender roles within same-sex couples (e.g., Sánchez et al. 2009), but the reasons and contexts for this type of satisfaction appraisal remain unknown. Future researchers should attend more specifically to the ways that gender roles and heteronormative assumptions across diverse relationship types and couples shape the ways that sexual pleasure and satisfaction are distributed between partners.

In a similar vein, narrative researchers have argued for a better understanding of the inter-subjective nature of the self. For example, Burkitt (2010) has argued,

[F]rom the earliest years our sense of self is intertwined with the voices of others…these voices can have their own autonomy, intruding into our self consciousness and our responses to others, often in unwanted, unplanned, unwilled, and surprising ways (p. 306).

This theoretical position presents a challenge to researchers working with data of all types to capture this complex web of self/other dynamics. While narrative researchers, like Burkitt (2010) and Raggatt (2010), provide important insights into the under-studied dynamic at play within individuals’ narratives, neither address the systematically gendered or heteronormative nature of this self/other dynamic. The coding and analysis procedures in the current study attended not only to how self/other dynamics appeared in participants’ narratives, but also to which others were embedded within the self-narratives to more closely examine additional patterns in these data.

Questions remain as to whether or not survey item can be written that can address the fluid subjectivity observed in this study. Certainly, many researchers have designed and tested questionnaires which include large numbers of items meant to test for various dimensions of sexual satisfaction (e.g., Hudson et al. 1981; Lawrance and Byers 1992; Meston and Trapnell 2005). Strongly worded instructions that direct the research participant to evaluate their own personal experience can be emphasized in survey design, which may diminish some of the patterns observed in this study. However, while these methodological steps would address aspects of this challenging psychological construct, we are still left with several important questions—regardless of our best attempts as researchers to forward the subjectivity of the participant: Whose satisfaction is imagined as possible, probable, and necessary? What are the gendered relational dynamics that distribute sexual satisfaction as a resource, and how does this affect the anticipation and demand for this resource (see McClelland and Opotow 2011)?

Satisfaction and Orgasm

Orgasm frequency and sexual satisfaction were moderately correlated in this study. While a statistical correlation found in this and other studies may demonstrate that orgasm consistency and satisfaction are positively related, the findings from the interview portion of the study shed additional light on this relationship by introducing questions such as: Whose orgasm correlates with whose satisfaction? How might orgasm be prioritized for some and not for others? In what types of relationships is orgasm considered important?

As this sample was comprised of mainly young adults, the relationships between satisfaction and orgasm may be different than in studies with older adults. For example, three (7%) of the women in this study had never experienced an orgasm during partnered sex or masturbation, which may diminish the association between orgasm and satisfaction for this subsample. However, when considering the rates of women of all ages who report never experience orgasm—ranging from 6.8% to 34%—this proportion is actually quite low (Graham 2010). However, as is evident in the enormous range observed in the rates of women who report never having an orgasm, researchers have emphasized different criteria for what “counts” as an orgasm during sexual activity (see Lloyd 2005) and it is difficult to know the conditions, relationships, and activities that predict how many women have never experienced orgasm.

Looking to related research with adolescents, Impett and Tolman (2006) found that 80% of the older adolescents (ages 16–19) reported that their most recent intercourse experience “made [them] happy,” while less than half (43%) reported that they “liked how [their] body felt” (p. 638). Thirteen percent of the sample reported no positive consequences from their most recent intercourse experience; in other words, they answered “no” to items that inquired whether it was a good experience, made them happy, made them feel closer to the other person, and whether they liked how their body felt. These findings suggest a high frequency of happiness, but a much lower frequency of positive embodied outcomes. This, in combination with many studies that demonstrate the tenuous association between female sexuality and pleasure (Fine and McClelland 2006, 2007; Tolman 1994, 2005), highlights how often young women’s sexual encounters occur in the absence of their own physical pleasure.

In the current study, lesbian and bisexual participants reported (non-significant, but trending) lower sexual satisfaction and orgasm rates as compared to the heterosexual participants. This finding may be evidence of what Horne and Zimmer-Gembeck (2006) found in their study of female sexual subjectivity and sexual orientation:

Compared to girls who identify as heterosexual, other girls (lesbians, bisexuals, heterosexuals with same-sex experience, and girls who were unsure) have a higher sense of entitlement to sexual pleasure from the self and from their partner, are more self-efficacious in achieving sexual pleasure, and reflect more on the sexual aspects of their life (p. 136).

If an LGBT identity, same-sex desire, and/or same-sex experience encouraged female participants to expect more from their sexual encounters, this shifting set of standards may explain the low satisfaction and orgasm rates reported by this group. Other possible explanations include that this group of lesbian, queer, and bisexual women experienced lower orgasm and satisfaction rates due to sexual stigma, partner choice, relationship quality, or a number of other factors.

Orgasm frequency was strongly correlated with liking sex with a partner. This may be because while orgasm and liking sex are both rooted in a specific set of experiences. Items that inquire about “liking sex with a partner” may provide useful insight into specific sexual encounters, encounters with a specific partner, or a range of other dimensions within this construct. This type of specificity may be helpful to researchers who are interested in collecting data on participants’ evaluations which do not require the participant to make the type of meta-level appraisals that are required within overall satisfaction appraisals.

Study Limitations

A number of design issues limit the generalizability of the findings described here. While the current study was able to demonstrate that there are differences in how individuals approach satisfaction appraisals, not enough is known about sexual expectations and how they vary by sexual minority status, age, gender, ethnicity, race, or immigration status—just to name a few.

Women were less defined by their sexual minority status in this study because there were very few identified lesbians in the sample and the bisexual women were mainly partnered with men at the time of the study. This may have reduced the social stigma related to sexual identity for the women in the study and it makes the heterosexual women and the bisexual women potentially more alike than they would be in other studies.

The current study is also limited by the young age of its sample. Previous research has shown that sexual concerns change over the life course (Carpenter 2010; Rossi 1994). As a result, it is unclear whether the findings here would generalize to men and women of all ages. However, as Carpenter et al. (2009) have argued, individuals’ expectations and experiences of sexual satisfaction may be more cohort-based than age-related. The young age of the participants in the study presents unique contributions and limitations to the study of sexual satisfaction. Young adults usually had fewer sexual partners than their older counterparts, which may provide them with fewer sexual experiences to use as comparisons when making satisfaction evaluations, thereby perhaps encouraging them to evaluate their current experiences as better or worse than they might later in life. Alternatively, this young age group also may have a greater sense of sexual possibility if they have not yet committed to a lifelong partnership and this may lead to a greater sense of control over their sexual partner choice and their sexual encounters as a result. Of course, this lack of stability may also introduce feelings of insecurity that encourage young adults to consider their partner’s experiences of satisfaction as relevant and primary due to the potential loss of that partner. This variety of contexts due to age, relationships status, access to sexual partners, cohort differences (just to name a few) are all important areas for future researchers to consider.

Implications for Research and Policy

The current study takes up the question of what it means to acknowledge the limits of what participants are able or want to express in research settings. It builds on earlier theoretical work in which Michelle Fine and I addressed the methodological dilemma of studying subjectivities when they are wrapped in the “cellophane” of social norms and public policies (McClelland and Fine 2008). The current study sits squarely within this same feminist methodological dilemma of documenting how an individual evaluates their sexual experience at the same time as “challenging women’s [and men’s] taken for granted experiences” (Kitzinger and Wilkinson 1997, p. 572). This dilemma becomes important when researchers report that individuals and groups report feeling satisfied, be it with their sex lives, their rates of orgasm, their relationships, etc., yet are also found to be using divergent criteria for these evaluations. This dilemma becomes especially salient—and uncomfortable—when research findings are translated quickly into clinical settings, diagnoses, and interventions which are not equipped to acknowledge these complexities.

For guidance, I turn to Josselson’s (2004) “hermeneutic of suspicion” which animates this methodological stance by requiring the researcher to approach self-report data with a “skepticism of the given” (p. 3). Building from psychoanalytic theories, Josselson’s attention to the manifest and the latent content of participants’ narratives signaled a turn in psychological methods when both became viable sites of knowledge (see also Kitzinger and Wilkinson 1997). Suspicion, in this framework, is not a route to truth, but a means to look beyond the face value of data in order to see what else is at play in the narrative. It is up to those of us working within this framework to develop methodological tools to carry out this work.

Implications for Self-Report Data

This study raises a difficult question: What is our responsibility as researchers when we find that participants provide self-report data without relying on their own subjectivity? Three possible (but certainly not exhaustive) responses are imagined. One, we might map this self/other narrative as further evidence of self-in-relation and develop theories and methods that can better capture this dynamic. A second alternative is to perhaps critique participants’ evident disavowal of subjectivity as not relational, but as a particularly gendered loss of self and agency. A third alternative response is to reverse our assumptions about sexual relationships more generally which implicitly insist on a narcissistic individual who “benefits” in some way from sexual encounters and instead, we might work to re-define the standard as one in which the other is the primary benefactor (thereby reversing the implicit pathologizing that is cast on those who rely on a partner’s appraisal).

These responses are not mutually exclusive, but are nevertheless in tension with one another. Each response prioritizes a different perspective, thereby making it a challenge to reconcile the three. Below, I examine these three responses as a way to begin to address the meta-level research issues raised by the findings presented in this study. While these issues of self and other might be more immediately relevant for researchers working on relationship and sexuality research questions, I argue that these issues are important to anyone working with quantitative or qualitative self-report data who must contend with the question of who the “self” is in self-report data.

Self-in-relation

As Raggatt (2010) has recently argued, the definition of a bounded self has been the assumed unit of analysis in psychology for the past 60 years:

From the 1950s, the humanistic and cognitive traditions revived interest in the self as a manifest concept, proposing models that emphasized containment and autonomy commensurate with the modernist worldview—in other words, a separation of outer from inner, subject from object, self from other, and public from private (p. 452).

Deconstructing the boundary between subject and object, sometimes referred to as theories of relational subjectivity, has captivated the social sciences for decades (Bakhtin 1981; Gilligan 1982; Winnicott 1953). This line of thinking has evolved tremendously with critical contributions from psychoanalytic theorists, psychologists, sociologists, and feminist scholars (McAdams 2001; Benjamin 1998; Drewery 2005), but at the heart of this work is the desire to theoretically describe the “intersubjective constitution of self-experience” (McAdams 2001, p. 343). In this framework, the individual is necessarily embedded within relationships with individuals, with groups, and with social norms and therefore should be studied not as separate, but in concert with this array of others.

The findings described in the current study may be evidence of the importance of relational aspects of sexuality for women in particular (e.g., Basson 2000; Kaschak and Tiefer 2001). Theoretical frameworks which emphasize the relational qualities of sexual development and experience offer a number of important contributions to research. For example, this perspective requires that sexuality researchers expand their unit of analysis from genital responsiveness, to the many other social, political, and relational aspects of sexual experiences. With this in mind, the findings described in this study potentially offer evidence of what this relational-ness looks like for some. The question remains whether this attention to relational subjectivity might obscure the mechanisms of specific relations, i.e., gender norms and heternormativity which may encourage specific forms of relationality over others, or perhaps, may stigmatize those individuals who do not desire to prioritize a relational subjectivity.

Disavowal of Subjectivity

An alternative reading of the results described in the current study focuses not on the gain implied by relational theories, but on a loss or disavowal of subjectivity in participants. When participants use an interpersonal benchmark such as their partner’s sexual satisfaction, this is conceptually different than research which has shown that sexual concerns are influenced by relational factors (Byers 2005; Sprecher 2002). The findings described here, rather than illustrating a relational dynamic, may be more aligned with a perspective where the person uses another’s satisfaction instead of their own. This seems to be less relational and perhaps more akin to substitution. There may be important costs associated with emphasizing relational qualities for women in particular; not all inter-subjective negotiations are dyadic in nature. Researchers must distinguish appraisals that are characterized as dyadic from appraisals where a participant substitutes another for oneself. These tensions and fusions among self, other, identity, and subjectivity are important to examine for their assumptions and gendered histories (Benjamin 1998; Chodorow 1978; Flax 1990; Layton 2002). As we move ahead with developing theories and methods that document the fluidity between these terms, it is essential that we not confuse subjectivity-in-relation with the under-development of subjectivity in the first place.

Partner as Primary Benefactor

Another possible response is to consider the implicit assumptions built into the fabric of our relational and sexual theories: that the individual should benefit from an encounter. Turning this model on its head, we might develop theories of sexual satisfaction where the partner is of primary concern instead of the self. This theoretical model would describe individuals as attending to intimate partner(s). Those who do not attend to partner(s) might, therefore, be interpreted as hyper-individualistic or narcissistic. This reversal of the traditional satisfaction model has the potential to reframe the relational perspective as less inherently feminine and instead, a model of high dyadic functioning. This reversal is the more radical and potentially most interesting because it reveals our assumptions about the modern self, sexual encounters, and who deserves to be rewarded.

In sum, there are a number of important and difficult questions raised by the findings for researchers, and specifically those researchers who rely on self-report data, to consider. First, what assumptions about unitary and self-enclosed “selves” are we making and how do these assumptions cause/allow us to miss essential aspects of the self-report process in research (see Danziger 1997)? Second, what are the gendered and heterosexist implications of this? In other words, do self-report measures “miss” or “misunderstand” some selves more than others? Are femininzed selves (or perhaps those selves that are encouraged strongly to attend to (male) others) more often overlooked in research? Third, what are the implications of these oversights?

Implications for Policy

On June 18, 2010, the US FDA’s Reproductive Health Drugs Advisory Committee rejected Boehringer Ingelheim’s petition for FDA approval of the drug flibanserin citing flibanserin’s clinical efficacy was “not robust enough to justify the risks” (Wilson 2010b). Much has been written about the potential implications of this (and other) so-called pink Viagra drugs (Wilson 2010a), which is unfortunately beyond the scope of this discussion. Critics have effectively argued that the larger umbrella of “sexual dysfunction” research should be examined for its assumptions and for its relationship to a pharmaceutical industry that has grown in response to newly defined sexual dysfunction category (Tiefer 2000). The recent example of flibanserin demonstrates the enormous weight given to sexual satisfaction in determining how medical and clinical professionals intervene in bodies depending on their evaluations of this complex psychological concept.

When researchers take the step to link their findings to the consequences their research could potentially have—on individuals, interventions, or policies—they are inevitably faced with a new set of challenges and responsibilities. In order to consider these consequences, a researcher must adequately theorize and capture potential sources of invalidity that might yield bias or unfairness for research participants, as well as those individuals affected by the research findings more broadly (Fernández-Ballesteros and Botella 2008, p. 103). While psychologists concerned with validity issues have long been interested in the impact of data outside of the laboratory (Cronbach and Meehl 1955; Messick 1980), this concept has been more recently named “impact validity”—a form of validity testing for researchers to evaluate the potential of research to affect policy (Barreras and Massey, forthcoming). At the heart of these theories linking research to consequences is the responsibility of making inferences from data and evaluating the “social considerations of the intended and unintended consequences of a measurement’s use” (Fernández-Ballesteros and Botella 2008, p. 103). When the data being evaluated are self-report data, these challenges are brought into sharp relief because this asks the researcher to critically analyze data about participants that were provided by participants.

The issue of bias in self-report data has been addressed by many psychologists over the years, with a particular focus on biases introduced by time (i.e., retrospective bias), social desirability, and question comprehension (see Schwarz 2007; Schwarz and Sudman 1994 for review). The current study introduces another form of potential bias in self-report data: implicit inclusion of the other in the self. Because this bias was observed within sexually marginalized groups (women and LGBT men), this question of bias is immediately linked with potential social and policy consequences. In sexual satisfaction research, if participants report on their partner’s level of satisfaction instead of their own, little will be understood about the psychology of these groups. A possible consequence is a new form of sexual marginalization—this time unintentionally articulated by the individual him or herself.

Conclusion

This study provides unusual insights into what people mean when they indicate that they are sexually satisfied. Findings indicated tremendous diversity in definitions and appraisal processes. Two groups emerged in the analysis: individuals who anchored satisfaction with the self and individuals who anchored it outside of the self. More specifically, this study found that it was women and LGBT men who, at times, anchored their satisfaction appraisal in a partner’s sexual satisfaction instead of their own.

These findings raise concerns about how gender norms and stigmatized sexual contexts shape expectations for sexual satisfaction and reveal how survey items may obscure individuals’ varying appraisal processes. With the increasing use of sexual satisfaction as an indicator of personal and relational health, critical evaluations of the cognitive processes that individuals use to make satisfaction appraisals are quickly becoming essential. This study directs researchers to include additional methods and measures that can illustrate the processes individuals are using to make evaluations concerning their own sexual health and to pay close attention to how participants articulate the inclusion of the other in the self when making satisfaction appraisals.