Measuring women’s sexual autonomy: Development and preliminary validation of the women’s sexual autonomy scale

Abstract

Background:

Sexual autonomy is an influential component of sexual health risk reduction frameworks, but a universal assessment of sexual autonomy is currently lacking.

Objectives:

This study develops and validates the Women’s Sexual Autonomy scale (WSA), a comprehensive measure that captures women’s perception of their sexual autonomy.

Design:

Forty-one items were initially created based on current research and in consultation with sexual health experts. In Phase I, a cross-sectional study with 127 women was conducted to finalize the scale. In Phase II, a cross-sectional study with 218 women was conducted to test the stability and validity of the scale. A confirmatory factor analysis was conducted with an independent sample of 218 participants.

Methods:

In Phase I, principal component analysis with promax rotation was conducted to examine the factor structure for the sexual autonomy scale. Cronbach’s alphas were conducted to assess the internal consistency of the sexual autonomy scale. In Phase II, confirmatory factor analyses were conducted to confirm the factor structure of the scale. Logistic and linear regressions were used to assess validity of the scale. Unwanted condomless sex and coercive sexual risk were used to test construct validity. Intimate partner violence was used to test predictive validity.

Results:

Exploratory factor analysis identified four factors across 17 items: 4 items on sexual cultural scripting (Factor 1), 5 items on sexual communication (Factor 2), 4 items on sexual empowerment (Factor 3), and 4 items on sexual assertiveness (Factor 4). Internal consistency for the total scale and subscales were adequate. The WSA scale showed construct validity by negatively relating to unwanted condomless sex and coercive sexual risk, and predictive validity by negatively relating to partner violence.

Conclusion:

The results of this study suggest the WSA scale provides a valid, reliable assessment of sexual autonomy for women. This measure can be incorporated into future studies investigating sexual health.

Introduction

Sexual autonomy is a potentially influential attribute of women’s sexual health. Broadly, sexual autonomy is the “human right to protect and maintain an informed decision over one’s body, one’s sexuality, and one’s sexual experience.” ¹ Sexual autonomy allows women to navigate sexual experiences, recognize their sexual feelings as distinct from societal pressures and desires, and exercise control in their own sexual decision-making. ² Maintaining a sense of sexual autonomy is critically needed for healthy sexual development, building respectful sexual relationships, and reducing one’s risk of poor sexual and reproductive health outcomes. ² For example, a longitudinal study among young heterosexual couples found that higher levels of women’s sexual autonomy protected against sexually transmitted infection (STI) acquisition. ³ A study among a nationally representative sample of married and partnered women found that high sexual autonomy was associated with greater use of modern contraception (e.g. condoms, oral contraceptives). ⁴ These studies suggest that interpersonal environments that allow women to make autonomous decisions regarding their sexual self can have significant positive impacts for women’s sexual and reproductive health.

To date, there is a scarcity of studies examining women’s sexual autonomy specifically, ⁴ but these few studies highlight the social, cultural, and economic factors shaping women’s ability to develop and exhibit sexual autonomy. Gendered sexual scripts may constrain women’s expression of their sexual self.^5,6 According to sexual scripting theory, ⁷ sexual preferences and behaviors are socially constructed based on societal expectations and gender roles. According to this theory, women are expected to adhere to a relationship-specific view of sexuality, which is characterized by stability, monogamy, and passiveness. ⁷ Sexually passive women may be reluctant to refuse sex ⁸ and other sexual practices if it is against their partner’s wishes. Furthermore, studies have found that sexually autonomous women have the ability to discuss and use condoms with their partner if they are interested in condoms. ⁹ Being able to communicate and assert her sexual wishes and interest in using contraception is a way for women to exhibit their sexual autonomy.^10,11 Contrarily, experiencing intimate partner violence (IPV) in a relationship could constrain women’s sexual autonomy.^3,9 Abusive partners may force and coerce women into sexual encounters against their will, which can reduce women’s sexual autonomy.^3,9 In addition, women may fear violent retaliation from their abusive partners if they discuss safer sex practices such as condom use.^12–14 In addition, education could help women develop and exhibit sexual autonomy. Women with formal education were more likely to exhibit sexual autonomy than women without formal education. ¹⁵ Creating opportunities for equitable access to sexual health education could provide important information to girls and empower them during their sexual development. ¹⁵

Despite the burgeoning research on women’s sexual autonomy, there is little to no consistent measurement or validation of this key construct. ⁴ This study demonstrates the development and preliminary validation of a multidimensional scale using two independent samples of U.S. sexually active women of reproductive age.

Conceptualizing sexual autonomy

A uniform definition for sexual autonomy is needed in order to accurately measure its constructs and empirically examine its relation to women’s sexual health outcomes. To date, definitions of sexual autonomy have been developed across disciplines including public health, psychology, and law. The World Health Organization ¹⁶ defines sexual autonomy as the ability and control to make decisions surrounding “one’s sexual life” in accordance with the individual’s personal and social ethics. Sexual autonomy among adolescents has been described as “knowing one’s sexual desire and pleasure, recognizing and discussing sexual wishes and boundaries, and learning to anticipate and prepare for sexual acts.” ² In psychology, sexual autonomy refers to “feeling that one’s actions in sexual contexts are freely chosen, authentic expressions of the self.”^6,17,18

Prior to the public health adoption, sexual autonomy was a concept argued within the confines of law and ethics as a human rights issue. In law, sexual autonomy is defined as the human and constitutional right for a free person to express their sexuality and sexual love in a way specific to the individual. ¹⁹ In the context of rape law, sexual autonomy is re-framed using feminist theory. Specifically, sexual autonomy is a “distinctive con of personhood and freedom” that must be protected against a spectrum of dehumanizing behaviors that attempt to limit sexual choices. ²⁰

Sexual autonomy is a related yet distinct construct from sexual agency. Sexual agency has been distinguished and examined within the disciplines of sociology, women and gender studies, and psychology. For example, Albanesi ²¹ defines sexual agency as the “willingness to exert power within a sexual encounter in an attempt to sway the outcome of events.” Similarly, Bay-Cheng ²² states that sexual agency is an individual’s effort to shape their immediate experiences or longer courses of life through sexuality. Thus, in contrast to sexual agency, sexual autonomy is focused on one’s ability to control and make decisions regarding their sexual life, not just their effort to do so. ²²

The current literature uses law, psychology, and public health to conceptualize sexual autonomy as a multidimensional construct. Specifically, we conceptualize sexual autonomy as a person’s right to freely express their sexual self. In this context, the sexual self is individualized and meaningful to a person and includes components such as sexual love, identity, wishes, boundaries, and desires. This definition of sexual autonomy aligns with the “autonomy as self-governance” conceptualization. ²³ This conceptualization describes a person who is aware of their goals, insightful, and well-balanced between dependence and independence. ²³ Therefore, a sexually autonomous person can recognize and communicate their sexual love, desires, pleasures, wishes, identity, and boundaries. This comprehensive definition of sexual autonomy aligns with some public health and psychosocial concepts. In particular, sexual autonomy comprises: (1) sexual communication, defined as negotiating sexual practices and discussing health information with intimate partners, ²⁴ (2) sexual assertiveness, defined as the right to control their body and no obligation to allow unwanted sexual experiences, ⁸ and (3) a lack of sexual passivity, defined as the reluctance to refuse unwanted sexual experiences. ¹⁸ Sexual autonomy also adds unique concepts such as one’s ability to express their sexual self outside the confines of cultural sexual scripts and norms.

In this study, our definition acknowledges sexual autonomy as a multidimensional concept similar but also distinct from other sexual psychosocial concepts. Given the complexity of sexual autonomy, it is important to develop and validate a reliable measure that assesses one’s sexual autonomy within these various domains.

To date, only two studies have developed an assessment for sexual autonomy. First, Sanchez et al. ²⁵ constructed a sexual autonomy latent variable (i.e. underlying, unobservable construct) by modifying three items from a general autonomy scale. ²⁶ These items were ranked on a 7-point Likert-type scale from 1 (not at all true) to 7 (very true). This study found important evidence that sexual submissiveness was negatively associated with sexual autonomy among undergraduate students in the United States. ⁶ A sexual autonomy latent variable is a novel approach; however, using three items may actually oversimplify the content and complexity of sexual autonomy.

Next, a sexual autonomy scale developed by Chang ²⁷ consisted of 13 items ranked on a 5-point Likert-type scale. For this scale, the definitions and items for sexual autonomy were specifically related to college students’ use of contraception and sexual behavior as opposed to one’s ability to express their sexual self freely. To date, there is no reliable, validated measure for sexual autonomy that addresses the multidimensionality of sexual autonomy from a lens that combines law, psychology, and public health, and has also been piloted among a diverse sample of women.

Current study

The purpose of the current study is to develop and validate a comprehensive measure that assesses sexual autonomy. This study consisted of two phases, with both phases being cross-sectional studies: (1) Phase I’s purpose was to develop a set of items based on conceptual knowledge and expert input that operationalizes sexual autonomy from a multidimensional lens to assess individuals’ ability to freely express their sexual self. In addition, Phase I was used to conduct an exploratory factor analysis and a psychometric evaluation to develop an initial sexual autonomy scale, and (2) Phase II included conducting a confirmatory factor analysis with a second sample and conduct additional tests of reliability and validity.

In Phase I, we conducted factor analyses to determine the appropriate factor structure for this scale. We also conducted a psychometric assessment of the scale as well as tests of preliminary validity. We hypothesized that four factors would be identified– sexual cultural scripting (Factor 1), sexual communication (Factor 2), sexual empowerment (Factor 3), and sexual assertiveness (Factor 4). We also hypothesized that the WSA scale would demonstrate good psychometrics including Cronbach’s alphas > 0.70. In Phase II, unadjusted and adjusted regression analyses were conducted to examine construct and predictive validity between the WSA scale and the validation measures (i.e. unwanted condomless sex, fearful to request safe sex practices, and intimate partner violence [IPV]). Next, for construct validity, we hypothesized that sexual autonomy would negatively relate with unwanted condomless sex and fear to request safe sex practices. For predictive validity, we hypothesized that IPV would negatively relate with sexual autonomy.

Phase I: scale development and exploratory factor analysis

During the first phase of the study, the initial pool of items was developed, and factor analyses were conducted to identify a psychometrically sound instrument.

Materials and methods

Procedures

In March 2016, 128 women participated in an online survey through Amazon’s Mechanical Turk (MTurk). MTurk is an online participant system, commonly used to conduct social science studies. ²⁸ Participants were asked to complete an online survey administered through Qualtrics. ²⁹ The study’s inclusion criteria were as follows: (1) aged 18 and older and (2) residing in the United States. The exclusion criteria for this phase were as follows: (1) non-English speaking and (2) inability to read and comprehend questions. Informed consent was obtained from each participant. At the end of the survey, participants were compensated $1 for participating in the study. Participants were also provided community resource information on mental health services. In Phase I, a power analysis indicated that a sample size of N = 100 was needed for linear regression to detect an effect size = 0.40 at a power of 0.8 and alpha of 0.05. The Yale University IRB approved all study procedures.

Participants

This sample identified as White (75.1%), Hispanic (8.7%), Black (6.6%), and Asian, Native Hawaiian or Other Pacific Islander, American Indian or Alaska Native (9.7%). The average age was 35.0 years (standard deviation [SD] = 11.8 years). Most of the sample (92.4%) had at least some college education. Most of the sample was either employed full-time (61.7%) or part-time (16.6%).

Item pool for sexual autonomy

An initial pool of 41 items were created from four a priori domains: (1) Freedom to express one’s sexual identity, (2) Freedom to express one’s sexual assertiveness, (3) Freedom to express one’s sexual permissiveness outside sexual scripts, and (4) Freedom to express one’s interest in safe-sex practices and preferences. These four domains were chosen based on previous research suggesting a significant association between sexual autonomy, sexual communication, sexual identity, refusal of sexual acts, discussing sexual practices, and sexual cultural scripts.^8,24,30 The content for the Sexual Autonomy Scale items were influenced by the following scales: Sexual Communication, ²⁴ Women’s Sexual Assertiveness, ⁸ Sexual Cultural Scripts, ³⁰ and Autonomy Scale. ²⁶ A 5-point Likert-type scale ranging from 1 (strongly agree) to 5 (strongly isagree) was created to capture item responses.

Next, three sexual health experts with extensive research in both women’s sexual health topics (e.g. sexual satisfaction and esteem, sexual practices, and communication) reviewed the initial pool of items for content and construct validity. Collectively, their feedback addressed: (1) length of the survey, (2) reading level, (3) order of the questions, (4) appropriateness of the 5-point Likert-type scale, and (5) if the items were relevant to our definition of sexual autonomy. Based on their feedback, condom use questions were clarified with “If I wanted to use a condom,” the number of double-barreled questions were reduced, and both positively and negatively worded items were included to reduce agreement bias.

After incorporating the experts’ feedback, the initial pool of sexual autonomy items was pilot-tested for interpretation and comprehension by 30 graduate-level students, and other academic researchers. Seventeen out of the 41 items were reversed coded. All the responses were summed to create a total score. Higher scores indicated more sexual autonomy.

Phase I statistical analysis

Descriptive statistics (frequencies, means) were conducted to describe characteristics of the sample, and to assess the normality of the scale variables. A principal component analysis with promax rotation was conducted to examine the factor structure for the sexual autonomy scale. An oblique rotation, specifically promax, was used because conceptually the underlying factors are correlated. A Kaiser-Meyer-Olkin score greater than 0.80 indicated that the data were appropriate for factor analysis. The scree plot was reviewed to determine the appropriate number of components. Cronbach’s alphas were calculated to determine the internal consistency of the overall scale and specific subscales. All descriptive and factors analyses were conducted using SPSS 21. ³¹

Results: scale construction and exploratory factor analysis

First, the item-total correlations were evaluated, and poorly performing items (correlations less than 0.40) were removed. As a result of this process, 13 items were removed, leaving a total of 28 items. Principal component analyses with promax rotation revealed a Kaiser-Meyer-Olkin measure of sample adequacy to be 0.81 (Table 1). Review of the scree plot suggest that four factors exist among the sexual autonomy items.

Table 1.

Factor loadings for the EFA and CFA analyses.

#	Factor 1: sexual cultural scripting	Phase I Sample	Phase II Sample
1	It is difficult for me to stay committed to one sexual partner at a time. [R]	0.75	0.80
2	I use my body and looks to get sexual attention from others. [R]	0.47	0.38
3	If a sexual partner sexually touches me at the wrong time, it’s my right to stop them.	0.67	0.70
4	I am able to turn down a chance to have sex with an attractive person, even if that person wants to have sex with me.	0.59	0.67
	Factor 2: sexual communication
1	Talking to sexual partners about what I like during sex is easy.	0.84	0.83
2	I feel embarrassed when I talk to sexual partners about my sexual desires. [R]	0.76	0.69
3	Sexual partners encourage me to talk about my sexual preferences.	0.72	0.63
4	If I wanted to, I would talk to a sexual partner about what they like sexually.	0.63	0.60
5	My sexual partners are open to my ideas about what we do sexually.	0.49	0.83
	Factor 3: sexual empowerment
1	Before I say my sexual preferences, I think about whether it will make a sexual partner upset.[R]	0.64	0.56
2	It would bother me if people didn’t like my sexual partners. [R]	0.61	0.36
3	In general, a sexual partner’s attitude about sexual preferences can affect my opinions about my own sexual preferences. [R]	0.55	0.47
4	I feel a sexual partner is not attracted to me if they say “no” to one of my sexual preferences. [R]	0.49	0.56
	Factor 4: sexual assertiveness
1	I would try to convince my sexual partners to use a condom even if it was not an easy task.	0.80	0.68
2	If I am not in the mood for sex, I will refuse to have sex with a sexual partner.	0.70	0.64
3	I am not concerned about hurting a partner’s feelings when I suggest we use condoms.	0.56	0.63
4	If I wanted to use a condom and a sexual partner refused to use a condom, I would not participate in sexual acts.	0.49	0.71

The four factors for the sexual autonomy scale are as follows: Factor 1: Sexual Cultural Scripting, Factor 2: Sexual Communication, Factor 3: Sexual Empowerment, and Factor 4: Sexual Assertiveness. Eleven items did not load on any of the factors, resulting in a final item pool of 17 using a cutoff of factor loadings greater than 0.40 ³² (Table 1). These four factors accounted for 43% of the variance explained. Factor 1, Sexual Cultural Scripting; contained four items; Factor 2, Sexual Communication, contained five items; Factor 3, Sexual Empowerment, contained four items; and Factor 4, Sexual Assertiveness, contained four items. The internal consistency for each subscale was adequate: alpha = .65 for Factor 1 Sexual Cultural Scripting, alpha = .80 for Factor 2 Sexual Communication, alpha = .63 for Factor 3 Sexual Empowerment, and alpha = .75 for Factor 4 Sexual Assertiveness.

Phase II: confirmatory factor analysis and validation

During the second phase of the study, the sexual autonomy scale was administered to an independent sample to test the stability of the four-factor structure and further test the validity.

Materials and methods

Procedures

From August 2017 to April 2018, 218 were recruited to participate in a research study focusing on relationships, sexual health, and biomedical HIV prevention use among women of reproductive age in Connecticut. The inclusion criteria for this phase were as follows: (1) cisgender women; (2) between the ages of 18 and 35; (3) reported at least one of the sexual risk indicators for pre-exposure prophylaxis (PrEP) candidacy in the past 6 months, according to the 2017 CDC clinical summary guidelines (i.e. unprotected sex with a male partner, HIV-positive sexual partner, recent STI, two or more sexual partners, transactional sex); (4) spoke English and/or Spanish; and (5) resided in Connecticut. The exclusion criteria for this phase were s follows: (1) non-English speaking, and (2) inability to read and comprehend questions.

We included the sexual autonomy scale as part of the assessment to conduct a confirmatory factor analysis and assess validity. Participant recruitment occurred online and throughout the community in Connecticut. Participants provided written informed consent and were able to complete the survey either in-person at the research office or online. At the end of the survey, participants were compensated $25 and provided a list of community resources. For Phase II, the power analysis indicated that a sample size of N = 208 was needed for a logistic regression model to detect an effect size = 1.50 at a power of 0.8 and alpha of 0.05. The Yale University IRB approved all study procedures.

Participants

This sample identified as White (43.4%), Hispanic (24.4%), Black (22.0%), and Asian, Native Hawaiian or Other Pacific Islander, American Indian or Alaska Native (10.2%). Within the sample, 36% had at least some college education, and 37.1% had completed college. Most of the sample was employed (60.5%).

Validation measures

Unwanted condomless sex

Participants were asked “How many times did you have sex without a condom when you wanted to use a condom in the past 6 months?” Participants were able to respond as: 0, 1, 2, or 3+ times. A summary, binary variable was created: unwanted condomless sex (reporting one or more times unwanted condomless sex occurred in the past 6 months). These items were developed by Teitelman, Tennille. ³³

Fear to request safer sex practices

Participants were asked to report if they experienced fear to discuss birth control, condoms, and refuse sex with an intimate partner in the past 6 months. The fear to request condom use and refuse sex were used from Decker, Miller. ³⁴ Three summary, binary variables were created: afraid to request condoms (reporting one or more times of being fearful to request condoms in the past 6 months), afraid to refuse sex (reporting one or more times of being fearful to refuse sex in the past 6 months), and afraid to discuss birth control (reporting one or more times of being fearful to discuss birth control in the past 6 months).

Physical intimate partner violence

Physical intimate partner violence (IPV) was assessed in the past 6 months using the physical assault subscale of the Conflict Tactics Scale-2 (CTS-2). ³⁵ Women were asked if they experienced physical forms of partner violence in the past 6 months by a romantic and/or sexual partner. Examples of physical partner violence includes “hitting, pushing, and/or being shoved.” A summary, binary variable was created: physical IPV (affirmative response to any form of physical IPV in the past 6 months versus no physical IPV in the past 6 months).

Sexual intimate partner violence

Sexual intimate partner violence (IPV) was assessed in the past 6 months using the Sexual Experiences Survey (SES). ³⁶ Women were asked if they experienced sexual forms of partner violence in the past 6 months by a romantic and/or sexual partner. Examples of sexual partner violence includes “kissed or touched in a sexual way when she did not want to.” A summary, binary variable was created: sexual IPV (affirmative response to any form of sexual IPV in the past six months vs no sexual IPV in the past 6 months).

Phase II statistical analysis

A second-order confirmatory factor analysis (i.e. items loading to subscales, subscales loading to sexual autonomy) was conducted using Mplus 8.5. to assess the stability of the four-factor structure. Given the potential multidimensional nature of construct, the errors of the four factors of the sexual autonomy scale were correlated. Standard goodness of fit statistics for confirmatory factor analysis was used to determine the model’s fit: factor loadings ⩾ 0.30; (1) a root mean square error of approximation (RMSEA) less than .08; and (2) a comparative fit index (CFI) and Tucker-Lewis index (TLI) ⩾ .90. Cronbach’s alphas were calculated to determine the internal consistency of the overall scale and specific subscales. Logistic regression analyses were performed to test associations between sexual autonomy (total and subscales) with unwanted condomless sex and coercive sex. Linear regression analyses were performed to test associations between sexual autonomy (total and subscales) with physical and sexual IPV victimization.

Results: confirmatory factor analysis and reliability

The model fit was strong: RMSEA = .05, CFI = .95, and TLI = .95. The factor loadings were moderate to high and ranged from .36 to .83. In the second-order model, the factor loadings were also moderate to high and ranged from .40 to .85.

Internal consistency was conducted to assess the reliability of the WSA scale and subscales (Table 2). The 17-item WSA total scale had good inter-item reliability with a Cronbach’s alpha of 0.80. The internal consistency for each subscale was adequate: alpha = .61 for Factor 1 Sexual Cultural Scripting; alpha = .81 for Factor 2 Sexual Communication; alpha = .60 for Factor 3 Sexual Empowerment; and alpha = .78 for Factor 4 Sexual Assertiveness.

Table 2.

Mean, standard deviations, and reliabilities of the total scale and subscales from Phase II.

	# of Items	M (SD)	Alpha
Total sexual autonomy	17	62.26 (9.73)	0.80
Sexual cultural scripting	4	15.72 (3.45)	0.61
Sexual communication	5	19.03 (4.29)	0.81
Sexual empowerment	4	13.35 (3.25)	0.60
Sexual assertiveness	4	15.06 (3.85)	0.78

M: means; SD: standard deviation; Alpha: Cronbach’s alpha.

The means and standard deviations for the total scale and subscales are shown in Table 2. For the total scale score, higher scores correspond to being sexually autonomous. For the individual subscales, higher scores corresponded to more adherence to that specific behavior. For example, responses to the Sexual Communication items were summed such that higher scores corresponded with more sexual communication.

Tests of validity

Construct validity was assessed with unwanted condomless sex (Table 3). Logistic regression analyses revealed that higher sexual autonomy (OR [95% CI] = .939 [.907, .972]), higher sexual communication (OR [95% CI] = .937 [.876, 1.003]), higher sexual empowerment (OR [95% CI] = .846 [.768, .932]), and higher sexual assertiveness (OR [95% CI] = .840 [.775, .911]) were significantly associated with lower odds of reporting unwanted condomless sex. Stronger adherence to sexual cultural scripting norms was associated with a higher odd of reporting unwanted condomless sex (OR [95% CI] = 1.237 [1.125, 1.360]).

Table 3.

Logistic regressions of sexual autonomy and subscales with sexual psychosocial outcomes from phase II.

Autonomy constructs	Unwanted condomless sex		Afraid to request condoms		Afraid to refuse sex		Afraid to discuss birth control
	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p
Total Sexual Autonomy	.939 (.907, .972)	.000	.909 (.858, .962)	.000	.919 (.877, .963)	.000	.920 (.871, .971)	.003
Sexual Cultural Scripting	1.237 (1.125, 1.360)	.000	1.808 (1.441, 2.268)	.000	1.434 (1.252, 1.642)	.000	1.763 (1.412, 2.200)	.000
Sexual Communication	.937 (.876, 1.003)	.060	.971 (.872, 1.081)	.59	.963 (.883, 1.050)	.38	.979 (.876, 1.093)	.70
Sexual Empowerment	.846 (.768, .932)	.001	.852 (.737, .985)	.030	.756 (.663, .863)	.000	.877 (.757, 1.015)	.07
Sexual Assertiveness	.840 (.775, .911)	.000	.679 (.585, .787)	.000	.770 (.693, .855)	.000	.709 (.616, .816)	.000

OR: odds ratio; CI: confidence interval.

Construct validity was also assessed with fear to request safer sex practices (Table 3). Logistic regression analyses revealed that higher sexual autonomy (OR [95% CI] = .909 [.858, .962]), higher sexual empowerment (OR [95% CI] = .852 [.737, .985]), and higher sexual assertiveness (OR [95% CI] = .679 [.585, .787]) were significantly associated with lower odds of reporting fear to request condoms. Stronger adherence to sexual cultural scripting norms was associated with higher odds of reporting fear to request condoms (OR [95% CI] = 1.808 [1.441, 2.268]).

Logistic regression analyses revealed that higher sexual autonomy (OR [95% CI] = .919 [.877, .963]), higher sexual empowerment (OR [95% CI] = .756 [.663, .863]), and higher sexual assertiveness (OR [95% CI] = .770 [.693, .855]) were significantly associated with lower odds of reporting fear to refuse sex. Stronger adherence to sexual cultural scripting norms was associated with higher odds of reporting fear to refuse sex (OR [95% CI] = 1.434 [1.252, 1.642]).

Logistic regression analyses revealed that higher sexual autonomy (OR [95% CI] = .920 [.871, .971]) and higher sexual assertiveness (OR [95% CI] = .709 [.616, .816]) were significantly associated with lower odds of reporting fear to discuss birth control. Stronger adherence to sexual cultural scripting norms was associated with higher odds of reporting fear to discuss birth control (OR [95% CI] = 1.763 [1.412, 2.200]). However, sexual communication was the only subscale that did not significantly relate to any of the fear to request safer sex practices measures.

Predictive validity was assessed with measures of physical and sexual IPV (Table 4). Linear regression analyses revealed that physical IPV was associated with lower sexual autonomy (B [95% CI] = -4.786 [-7.523, -2.048]), lower sexual communication (B [95% CI] = -2.419 [-3.621, -1.216]), and lower sexual assertiveness (B [95% CI] = -2.017 [-3.100, -.933]). Physical IPV was also associated with stronger adherence to sexual cultural scripting norms (B [95% CI] = 1.740 [.821, 2.659]). Linear regression analyses also revealed that sexual IPV was associated with lower sexual autonomy (B [95% CI] = -6.264 [-9.308, -3.221]), lower sexual communication (B [95% CI] = -1.396 [-2.777, -.014]), lower sexual empowerment (B [95% CI] = -1.648 [-2.681, -.615]), and lower sexual assertiveness (B [95% CI] = -3.029 [-4.209, -1.848]). Sexual IPV was also associated with stronger adherence to sexual cultural scripting norms (B [95% CI] = 3.001 [2.006, 3.995]).

Table 4.

Linear regressions of sexual autonomy and subscales with intimate partner violence from phase II.

Predictors	Total sexual autonomy	Sexual cultural scripting	Sexual communication	Sexual empowerment	Sexual assertiveness
	B (95% CI)	B (95% CI)	B (95% CI)	B (95% CI)	B (95% CI)
Physical IPV	-4.786 (-7.523, -2.048)	1.740 (.821, 2.659)	-2.419 (-3.621, -1.216)	-.646 (-1.576, .284)	-2.017 (-3.100, -.933)
p	.001	.000	.000	.17	.000
Sexual IPV	-6.264 (-9.308, -3.221)	3.001 (2.006, 3.995)	-1.396 (-2.777, -.014)	-1.648 (-2.681, -.615)	-3.029 (-4.209, -1.848)
p	.000	.000	.048	.002	.000

CI: confidence interval; IPV: intimate partner violence.

Discussion

The Women’s Sexual Autonomy (WSA) scale is a multidimensional measure of sexual autonomy with demonstrated reliability, and construct and predictive validity. The current study hypothesized a priori that four subscales would emerge. Our hypothesis was supported as four a priori subscales did emerge: Sexual Cultural Scripting, Sexual Communication, Sexual Empowerment and Sexual Assertiveness. Together, these four subscales adequately captured the complexity of sexual autonomy: Sexual Cultural Scripting, Sexual Communication, Sexual Empowerment and Sexual Assertiveness. Conceptually, each of these subscales are demonstrating different but related dimensions of women’s sexual autonomy.

Preliminary tests of the WSA scale suggest it is a reliable and valid instrument, and its development expands current research on sexual autonomy in three ways. First, the present study examined sexual autonomy among women and non-college samples. While most of the derivation sample was college-educated, the generalizability of the scale was enhanced because the factor structure and validity of the scale was confirmed on a second sample where only approximately one-third were college educated, and there was more variability in racial and ethnic identity. Second, the WSA scale underscores a multidimensional construct of sexual autonomy compared to previous studies. The WSA scale encompasses an interdisciplinary perspective of sexual autonomy that draws from law, psychology, and public health. Therefore, creating a conceptualization and assessment of sexual autonomy that builds from these three distinct areas allows for an in-depth examination of women’s sexual autonomy across multiple domains. Finally, our WSA scale expands previous examinations of sexual autonomy by adding a unique element: sexual scripting. It is possible that a women’s sexual autonomy may be defined or largely influenced by sexual scripts in heterosexual relationships. The WSA incorporates items that relate to women’s ability to express their sexual autonomy outside of the confines of socially constructed sexual behavior in intimate heterosexual relationships.

Furthermore, the WSA scale demonstrated construct and predictive validity. Our findings indicate that higher sexual autonomy relates to lower odds of unwanted condomless sex and fear to request safer sex practices in the past 6 months. Conceptually, these relationships were hypothesized because women with constrained opportunities to freely express their sexual self may be placed in difficult sexual situations such as being afraid to discuss safe sex practices and engaging in condomless sex against their will. Similarly, our study also found that women who experienced physical and sexual IPV were more likely to report lower sexual autonomy. Consistent with existing research on IPV and sexual autonomy, ³ these findings underscore how relationship dynamics, in particular partner behaviors, can affect women’s sexual autonomy. Physically and sexually abusive partners may use controlling tactics to constrain women’s ability to express their sexual self. Women experiencing IPV may also feel a lack of control over sexual decision-making in the relationship, especially if there is concern about violent retaliation from a partner.

In addition to the relationship of overall sexual autonomy with sexual health outcomes, the sexual autonomy subscales also had interesting and informative relationships with the outcomes. In general, sexual autonomy subscales such as sexual communication, sexual empowerment, and sexual assertiveness were protective against unwanted condomless sex, and fearful request of safe sex practices. These subscales were higher, on average, among women without IPV experiences. However, stronger adherence to the sexual cultural scripting subscale related to reports of unwanted condomless sex, fearful request of safe sex practices, and were higher, on average, among women experiencing IPV.

Sexual scripting theory may offer some potential explanations for these findings. First, women who strongly believe in sexual cultural scripts may be more at risk for unwanted condomless sex and more fearful to request safe sex practices. Coupled with our findings, some women may feel compelled by sexual cultural scripts to engage in sex to maintain stability in their relationship. Furthermore, existing research indicates a significant relationship between IPV and sexual cultural scripts^37–39 such that sexual cultural scripts may legitimize men’s use of IPV against women. Therefore, it is possible sexual cultural scripts are introduced and/or reinforced in the relationship when men use of IPV against women.

These findings should be interpreted in the light of the following limitations. Due to the sensitive nature of the survey questions (i.e. sexual activity, IPV), these responses are susceptible to social desirability bias, such that women may have responded in a way that is “favorable” to others. Our participants from our first phase were recruited from MTurk and the northeastern region of the United States for the second phase. Therefore, the generalizability of our findings is limited, and future studies with a larger pool of participants are needed to replicate these findings in more diverse populations. While the overall scale demonstrated good consistency, two subscales (i.e. Sexual Cultural Scripting and Sexual Empowerment) had low consistencies. The factor loadings for both subscales were moderate in each phase, and in Phase II, both subscales loaded on the second-order model. Given the strength of the factor analyses in addition to their unique dimensions of sexual autonomy, we decided to retain both subscales in the final scale development phase. Sexual health is still considered a sensitive topic, ⁴⁰ and while the overall scale addresses sexual autonomy, it is possible that sexual scripting and empowerment may be difficult topics for women to agree on how to express unlike the well-performing subscales of Sexual Communication and Assertiveness (i.e. sexual scripting and empowerment may be expressed differently among women). Future research could conduct an in-depth qualitative analysis of women’s expression of sexual cultural scripting and empowerment in order to refine the existing subscales. This scale was constructed for cisgender women engaged in sex with male partners which allow an in-depth investigation into gender roles for heterosexual cisgender women. This scale may not represent the sexual autonomy for cisgender men, and transgender women and men. This study was unable to include a measure of test–retest reliability to strengthen the stability of the instrument over time. Future implementation and refinement of the WSA scale should examine measurement invariance across racial and cultural groups of women (e.g. Black and Latina women). Future refinement of the WSA scale should also include other validation measures capturing sexual love, desires, satisfaction, and sexual scripts. Future studies should also consider the implications of women’s sexual autonomy in romantic couples or dyads.

Conclusion

This is one of the first reliable and validated tools that assessed sexual autonomy among women in the United States. This study adds to the nascent body of literature addressing sexual autonomy among women. The WSA scale provides an opportunity for future research to understand and highlight a currently understudied component of sexual health. While sexual autonomy has been emphasized and addressed in qualitative studies, the WSA scale can empirically demonstrate the importance of sexual autonomy and other sexual health outcomes including sexually transmitted infections and HIV prevention methods.