Pornography,provocative sexual media,and their differing associations with multiple aspects of sexual satisfaction

Theoretical background

Associations between the viewing of sexual content and sexual satisfaction have been suggested by several theoretical frameworks, including script theory (Gagnon & Simon, 1973) and objectification theory (Fredrickson & Roberts, 1997). Sexual script theory claims that sexuality is socially constructed by values, personal experiences, and the mass media (Gagnon, 1990). These scripts may be especially meaningful for heterosexual couples, as many messages in sexual content present a heteronormative script, with men’s sexual dominance and women’s subservience (Bridges, Wosnitzer, Scharrer, Sun, & Liberman, 2010). Recent longitudinal research has suggested that sexual content has a directional influence on sexual attitudes and behavior (Braithwaite, Coulson, Keddington, & Fincham, 2015). Stulhofer, Busko, and Landripet (2010) have also empirically validated the theoretical assertion of sexual content influencing sexual satisfaction through a number of mechanisms.

Objectification theory (Fredrickson & Roberts, 1997) has the potential to complement sexual script theory as it could be argued that some sexual content depicts a script of objectification (Klaassen & Peter, 2015). According to objectification theory, the Western sociocultural emphasis on women’s beauty can result in men objectifying women and women self-objectifying. One study that coded for the level of objectification found in participants’ favorite types of media content (television shows, films, and magazines) showed that men’s use of sexual content was associated with lower sexual satisfaction, partially because they were more likely to objectify their partner (Zurbriggen , Ramsey, & Jaworski, 2011). Meanwhile, women’s use of sexual content has been found to be associated with higher self-objectification (Vandenbosch & Eggermont, 2012; Zurbriggen et al., 2011), which is associated with lower sexual satisfaction (Claudat & Warren, 2014).

Multiple aspects of sexual content and sexual satisfaction

Despite clear theoretical and limited empirical evidence of links between viewing sexual content and sexual satisfaction, such research to date has a major limitation: the almost universal operationalization of sexual content viewed and sexual satisfaction as monolithic constructs. Malamuth, Addison, and Koss (2000) noted the difficulties in operationalizing use of sexual content, as questions typically used continue to be vague and vary from study to study (Short, Black, Smith, Wetterneck, & Wells, 2012). For example, many studies include general questions about how often participants have used sexual content over an allotted period of time, without any specification of what type of sexual content was used (Short et al., 2012). In fact, researchers also have often neglected defining what they mean by “pornography,” not letting participants know whether they are referring to explicit sexual acts, nude pictures, or other material.

Noticing the consistent one-item measurement and occasionally eclectic methodology used in assessing use of sexual content, studies by Hald and Stulhofer (2016) and Willoughby and Busby (2016) have suggested perhaps the clearest call for a multidimensional approach to sexual content viewed. Hald and Stulhofer (2016) showed that paraphilic (e.g., violent/coercive) pornography differed from mainstream pornography (e.g., vaginal or oral sex). Meanwhile, Willoughby and Busby (2016) found variation in what content participants considered pornographic, with two clear clusters of perceptions emerging. Within their results, they found that the vast majority of participants agreed upon certain forms of sexual content being defined as pornography (e.g., “A video that graphically depicts a three-way sexual encounter”), while other sexual content in media was not considered to be pornographic (e.g., “An image of a woman alone posing in a suggestive way with underwear on.”), suggesting at least a possible clear distinction between sexual content depicting explicit sexual acts and more general, provocative sexual media that is not explicit.

Like sexual content, scholars have also suggested that sexual satisfaction may be best represented as multifaceted. McClelland (2010, 2011; McClelland & Hunter, 2013) is one scholar in particular who has theoretically interrogated what it means to be sexually satisfied and suggested that scholars’ measurement does not fully capture its multifaceted nature and complexity. One weakness with the measurement of sexual satisfaction is the use of vague, global definitions assessing the subjective assessment of one’s sexual relationship (McClelland, 2010; Sprecher & Cate, 2004). Pascoal et al. (2014) added their voices to the critiques of sexual satisfaction measurement by pointing out that the research has suffered from scarce conceptual definitions of sexual satisfaction, few theoretical models of the understanding of sexual satisfaction, and inconsistency in the indicators used in different measures of sexual satisfaction.

Noticing this gap in the literature surrounding sexual satisfaction measurement, Pascaol et al. (2014) asked participants open-ended questions about what constitutes sexual satisfaction and found several distinct constructs in assessing sexual satisfaction: ludic sexuality (e.g., variety, creativity), sexual frequency, and sexual intimacy (e.g., love and affection, expression of feelings, romance). Additional factors that have been suggested as being worthy of consideration are satisfaction with time spent on foreplay and the amount of time spent on intercourse itself (Rust & Golombok, 1985). Ludic sexuality (Hally & Pollack, 1993), sexual frequency (McNulty, & Fisher, 2008), sexual intimacy (Bois et al., 2016), and satisfaction with time spent on foreplay and intercourse (Rust & Golombok, 1985) have all been evaluated as important indicators for sexual satisfaction and the overall health of the couple relationship. But no study to our knowledge has attempted to evaluate all of these constructs together within the context of viewing sexual content.

Current study

Because different types of sexual content may differ in the type of sexual script being presented, it is likely important to differentiate the type of sexual content being portrayed when exploring associations with sexual satisfaction. We suspected that explicit pornographic content would differ from provocative sexual media in its influences on sexual satisfaction because pornography and provocative sexual media present different sexual scripts. In our study, we evaluated how both pornography use and provocative sexual media use are associated with several aspects of sexual satisfaction that have been suggested in previous research (Pascoal et al., 2014; Rust & Golombok, 1985): time spent on foreplay, variety, the overall sexual relationship, love and affection, time spent on intercourse, and frequency.

Considering meta-analyses showing solitary sexual behavior to be the biggest distinguisher of gendered sexual practices (Oliver & Hyde, 1993; Petersen & Hyde, 2010), we also decided to evaluate all associations separately by gender. Sexual script theory suggests the importance of evaluating these associations for each gender, as pornography’s content (Bridges et al., 2010) presents different sexual scripts for each gender. Therefore, sexual expectations formed from viewing pornography may influence sexual satisfaction differently for men and women. Also, objectification theory suggests the value in assessing these associations for each gender, as the difference between men objectifying women and women self-objectifying may result in differing outcomes. With support from our chosen theories, we asked the broad question (RQ1), how are increased use of pornography and provocative sexual media associated with several aspects of sexual satisfaction for both men and women?

Sample and procedure

The sample of the present study consisted of 858 individuals in relationships, sampled from the Amazon Mechanical Turk (MTurk) website (http://www.mturk.com). The criterion for being in a committed relationship was self-defined, as we asked participants “Are you currently in a committed romantic relationship of any kind?” MTurk is a website dedicated to online labor and is used to employ “workers” from around the world to complete specific tasks. In regard to the current project, a job was posted inviting participants to complete a short 10- to 15-min survey on dating and relationships. Participants were instructed that they needed to be English speaking, and those who were interested were directed to a separate website to complete an online assessment. Before beginning the online survey, participants were asked to indicate consent and were informed about their rights as a research participant. Upon completion of the survey, participants were thanked for their time and given compensation of US$0.25, a rate similar to comparable tasks on the MTurk website. Scholars have noted that samples from MTurk are very similar to other diverse behavioral research samples, and several scholars have replicated previous research results on MTurk to provide evidence for the validity of the sample (Paolacci, Chandler, & Ipeirotis, 2010; Suri & Watts, 2011). All demographic information is included in Table 1.

OPEN IN VIEWER

Table 1.

Descriptive statistics for demographic control variables.

Variables	%	M	SD	Range
Age		33.10	9.36	18–77
Gender
Men	53.50
Women	46.50
Sexual orientation
Heterosexual	89.98
Sexual minority	10.02
Race
White	61.31
Asian	25.87
African	3.73
Latino	3.73
Mixed/Biracial	2.91
Native American	2.10
Other	0.35
Relationship status
First marriage	54.08
Cohabiting	17.37
Never married	16.43
Other	12.12
Education
High school or less	9.56
Some college	34.27
4-Year degree	39.39
Advanced degree	16.79
Religious affiliation
Conservative Christian	15.73
Roman Catholic	13.64
Hindu	11.42
Liberal Christian	13.29
Atheist	10.37
Agnostic	8.51
No affiliation	16.55
Other	10.49
Residency
U.S.	68.38
International	31.62
Employment
Full-time	59.96
Part-time	36.09
Unemployed	3.96
Number of children
0	43.36
1	24.83
2+	31.81

We increased confidence in the quality of the responses by carefully assessing how long each participant spent on the survey and having several reading comprehension questions. In our questionnaire, we had three Likert scale–type questions stating things like, “If you are reading this, select ‘3 to 5 days a week’.” If they filled in any option that was not “3 to 5 days a week,” we did not use their data. Aside from the reading comprehension questions, we also eliminated participants if they took far less than the allotted period of time (<6 min). For those who took slightly less than 10 min (6–10), we closely inspected their answers for any major inconsistencies that suggest little thought in completing the survey. After careful evaluation, 304 individuals were dropped from the original 1162 participants who were in a committed relationship to end up with our final sample of 858 individuals. A multivariate analysis of variance (MANOVA) was also conducted to assess the demographic differences between those who completed the survey adequately as opposed to those who incorrectly answered security check questions. A significant difference was found in the MANOVA, Wilks λ = .917, F (6, 1028) = 15.59, p < .01. Participants who completed the survey adequately were more likely to be White (p < .01), heterosexual (p = .03), less religious (p < .01), and have no children (p < .01). This could suggest some potential bias in participants that should be considered with interpreted results.

Measures

Measurement for both pornography use and provocative sexual media use was derived from a study by Willoughby and Busby (2016). In the study, participants were given 20 items and asked to rate the extent they believed each item to be pornography (0 = definitely not pornography; 10 = definitely pornography). In our study, the items used for pornography were those most agreed upon in the study of Willoughby and Busby (2016) to be pornographic content (all had a mean above 8.0). Meanwhile, the items used for sexual media were items with lowest consensus from the study of Willoughby and Busby (2016) about the content being pornographic (had a mean rating lower than 5.0). Again, the biggest distinction between these two aspects of sexual content was that the pornography scale assessed explicit sexual acts, whereas the provocative sexual media scale assessed suggestive, nonexplicit sexual media. Of note, we specifically focused on the provocative portrayal of women in sexual media, as research suggests it to be a very prevalent, specific subtype of sexual media. In fact, studies have suggested that women are sexually objectified in 52% of magazine advertisements (Stankiewicz & Rosselli, 2008). We were concerned that if we created a scale focused on the provocative portrayal of both men and women, then it would lack internal consistency and the results would be less interpretable.

Descriptive statistics

Gender differences were explored for all the variables. A MANOVA revealed an overall significant gender difference, F (17, 807) = 14.73, p < .01, partial η² = .24. Levene’s test was significant for several variables (all p < .01), including race, sexual orientation, U.S. residence, whether the participant has a child, pornography use, sexual media use, and sexual satisfaction with time spent on foreplay. This suggests that we should be cautious in directly comparing pornography and sexual media use of men and women. Table 2 shows means, standard deviations, ranges, and F values for each variable.

OPEN IN VIEWER

Table 2.

Descriptive statistics and gender differences of control, independent, and dependent variables.

	Men		Women		Response range
Variable	M	SD	M	SD		F value
Age	32.99	9.37	33.11	9.73	18–77	0.006
Race	0.42	0.49	0.36	0.48	0–1	2.75
Sexual orientation	0.08	0.27	0.13	0.33	0–1	6.01*
Education	4.51	1.27	4.11	1.28	1–7	20.71***
Impulsivity	2.21	0.93	2.25	0.92	1–5	0.48
Body image	3.40	0.80	3.12	0.89	1–5	12.37***
Religiosity	3.07	0.99	3.01	1.02	1–5	0.592
U.S. Resident	0.38	0.49	0.25	0.44	0–1	16.54***
Parent	0.52	0.50	0.59	0.49	0–1	4.58*
Pornography use	2.64	1.13	1.80	0.97	1–6	131.67***
Sexual media use	2.06	1.09	1.57	0.86	1–6	54.31***
Foreplay time	3.51	1.04	3.49	1.23	1–5	0.053
Variety	3.44	1.26	3.79	1.23	1–5	16.59***
Overall	3.80	1.07	3.95	1.07	1–5	3.88*
Frequency	3.12	1.17	3.50	1.14	1–5	15.04***
Love/Affection	3.82	1.24	3.94	1.23	1–5	1.66
Intercourse time	3.56	1.21	3.81	1.12	1–5	8.47**

*p < .05; **p < .01; *** p < .001.

Bivariate associations

Table 3 shows correlations between all variables in the study, but we briefly report the highlights from our variables of interest. For men, greater pornography use was associated with less satisfaction with sexual variety (r = −.11, p = .02), love and affection (r = −13, p < .01), and time spent on intercourse (r = −.21, p < .01) but positively associated with satisfaction with frequency (r = .10, p = .04). Men’s provocative sexual media use was associated with lower sexual satisfaction with love and affection (r = −.19, p < .01) and time spent on intercourse (r = −.17, p < .01) but higher satisfaction with time spent on foreplay (r = .14, p = .03) and frequency (r = .24, p < .01). For women, pornography use was not associated with any of the sexual satisfaction outcomes. However, provocative sexual media use was associated with lower sexual satisfaction with love and affection (r = −.12, p = .02) and time spent on intercourse (r = −.11, p = .04).

OPEN IN VIEWER

Table 3.

Bivariate correlations between control variables, independent variables, and outcome variables.

	Variable	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17
1.	Age	—	−.17*	−.11*	−.03	−.10*	−.05	.08	−.11*	.40*	−.18*	−.06	−.06	−.08	−.10	−.08	−.01	−.06
2.	Race	−.16*	—	—	.23*	.11*	.17*	.20*	—	—	.24*	.22*	−.03	−.00	−.10	−.06	−.07	−.02
3.	Sexual orientation	.01	—	—	−.06	.03	.02	−.14*	—	—	.13*	.07	.09	−.04	.04	−.02	−.02	.03
4.	Education	.09	.25*	.02	—	−.13*	.20*	.08	.34*	−.04	.13*	.20*	.11*	.11*	.08	.12*	.14*	.11*
5.	Impulsivity	.08	.19*	.09	.02	—	−.34*	.06	.07	.03	.21*	.14*	−.22*	−.21*	.10	−.22*	.24*	.22*
6.	Body image	−.01	.21*	−.089	.07	−.23*	—	.12*	.16*	−.02	.11*	.10*	.23*	.23*	.22*	.29*	.29*	.16*
7.	Religiosity	−.02	.29*	−.08	.15*	.05	.21*	—	.14*	.21*	−.05	.09	.13*	.05	.03	.10	−.01	.10
8.	U.S. Residency	−.21*	—	—	.31*	.26	.20*	.27*	—	.09	.30*	.35*	.09	.12*	−.01	.09	−.06	.04
9.	Parent	.36*	—	—	.08	.06	.05	.18*	.06	—	−.08	−.04	−.07	−.10	−.08	−.05	−.19*	−.03
10.	Pornography use	−.20*	.11*	.15*	.01	.25*	−.04	.16*	.13*	−.06	—	.68*	−.01	−.02	−.04	.03	−.07	−.09
11.	Sexual media use	−.22*	.39*	−.04	.15*	.33*	.09	.13*	.44*	.08	.59*	—	.01	−.10	−.08	.03	−.12*	−.11*
12.	Foreplay time	−.17*	.10*	−.06	−.01	−.09	.27*	.12*	.13*	−.07	.04	.14*	—	.43*	.55*	.56*	.40*	.34*
13.	Variety	−.18*	.08	−.06	−.05	−.19*	.25*	−.02	.10*	−.04	−.11	.05	.30*	—	.50*	.51*	.58*	.47*
14.	Overall	−.24*	.08	−.09	.03	−.09	.26*	.15*	.15*	.11*	.02	.07	.60*	.35*	—	.57*	.55*	.39*
15.	Frequency	−.32*	.21*	−.07	.01	.01	.34*	.17*	.30*	−.07	.10*	.24*	.52*	.28*	.58*	—	.47*	.38*
16.	Love/Affection	−.11*	−.00	−.06	−.04	−.28*	.19*	−.01	.03	−.13*	−.13*	−.19*	.32*	.50*	.46*	.32*	—	.42*
17.	Intercourse time	−.12*	.01	−.11*	−.10*	−.26*	.20*	.00	−.02	−.08	−.21*	−.17*	.21*	.49*	.31*	.24*	.55*	—

Note. Men’s correlations are below the line. Women’s correlations are above the line.

*p < .05.

Structural equation models

For each structural equation model, a comparative fit index (CFI) of .95, a root mean square error of approximation (RMSEA) of .08, and a standardized root mean square residual (SRMR) of .08 as they represent conservative estimates (Browne & Cudeck, 1993; Kline, 2010). All analyses were conducted in MPlus 7.0 using robust maximum likelihood to account for nonnormality in the data. There were 23 participants who had some level of missingness (<3% of the data), which were accounted by using maximum likelihood. We first conducted a confirmatory factor analysis and measurement invariance tests to better understand the psychometric properties of our sexual content constructs. Afterward, we created measurement models for our latent variables and full structural models to test our hypotheses.

Measurement model for women

Modification indices suggested multiple error correlations for the latent construct of pornography use and sexual media that substantially improved model fit. The final measurement model fits the data adequately (χ²(106) = 188.13, p < .01, CFI = .98, RMSEA =.05, SRMR = .04), with all factor loadings above .54 (see Table 4 for factor loadings of all latent variables for both genders).

OPEN IN VIEWER

Table 4.

Factor loadings for all latent variables.

Items	Men	Women
Pornography use
An image of a heterosexual couple having sex, which shows the man’s penis penetrating the woman.	.74	.79
A video showing two naked women or men manually stimulating each other.	.83	.88
A video of a woman or man alone masturbating.	.90	.89
A video that graphically depicts a three-way sexual encounter.	.73	.89
Sexual media use
A swimsuit issue of a magazine showing models in skimpy swimsuits or strategically covered topless views in a variety of provocative poses.	.88	.73
A television program showing the filming of a swimsuit edition of a magazine that shows models being filmed in a variety of provocative poses but no full nudity of breasts or genitalia.	.88	.80
An image of a woman alone posing in a suggestive way with underwear on.	.89	.87
A television show focused on strippers includes multiple instances of blurred nudity.	.78	.85
Impulsivity
I have difficulty controlling my temper.	.82	.80
I get so frustrated I am ready to explode.	.86	.88
I get upset easily.	.79	.78
Often I am afraid I will lose control over my feelings.	.81	.81
I slam doors when I am mad.	.67	.54
Body image
I’m looking as nice as I’d like to.	.82	.86
I worry about the way I look (reverse coded).	.48	.61
I’m pretty happy about the way I look.	.84	.89
I like what I see when I look in the mirror.	.82	.78

Structural model for men

The model showed adequate fit (χ²(274) = 488.43, CFI = .96, p < .01, RMSEA =.04, SRMR = .04) and explained the following variances for the different dimensions of sexual satisfaction: variety (R² = .15), overall (R² = .15), love and affection (R² = .17), time spent on foreplay (R² = .12), frequency (R² = .25), and time spent on intercourse (R² = .14).

Figure 1 shows that higher pornography use among men was significantly associated with lower satisfaction with sexual variety (β = −.16, p = .01) and lower satisfaction with the amount of time spent on intercourse (β = −.15, p = .02). Also, greater use of sexual media for men was significantly associated with lower satisfaction with the love and affection in the sexual relationship (β = −.20, p < .01).

OPEN IN VIEWER

Figure 1.

Structural equation model for men’s use of sexual content and sexual satisfaction. χ²(274) = 488.43, p < .01, CFI = .98, RMSEA = .04, SRMR = .04. Analyses controlled for age, race, sexual orientation, education, impulsivity, body image, religiosity, whether the participant had a child, and U.S. residency. Space constraints required abbreviations of several terms. “Foreplay” indicates satisfaction with time spent on foreplay. “Love” indicates satisfaction with love and affection. “Intercourse” indicates satisfaction with time spent on intercourse. *p < .05.

For the control variables, higher impulsivity was associated with lower satisfaction with variety (β = −.19, p = .01), love and affection (β = −.24, p < .01), and time spent on intercourse (β = −.20, p < .01). Higher body image was associated with higher sexual satisfaction with time spent on intercourse (β = .12, p = .03), love and affection (β = 14, p = .02), frequency (β = .33, p < .01), overall (β = .26, p < .01), variety (β = .18, p = .01), and time spent on foreplay (β = .24, p < .01). Higher age was associated with lower sexual satisfaction with time spent on foreplay (β = −.11, p = .02), variety (β = −.18, p < .01), overall (β = −.18, p < .01), frequency (β = −.19, p < .01), and time spent on intercourse (β = −.12, p = .02). Additionally, religiosity was associated with lower satisfaction with sexual variety (β = −.13, p = .02). Finally, those who were not residents of the U.S. reported higher satisfaction with the frequency (β = 20, p < .01) and love and affection (β = .20, p < .01) in their sexual relationships. Whether men had kids, race, sexual orientation, and education did not significantly predict any of the dependent variables.

Structural model for women

The overall model again showed adequate fit (χ²(274) = 387.94, p < .01, CFI = .98, RMSEA = .03, SRMR = .03) and explained the following variances of sexual satisfaction: variety (R² = .14), overall (R² = .11), love and affection (R² = .18), time spent on foreplay (R² = .13), frequency (R² = .15), and time spent on intercourse (R² = .10). Upon finding that the model fit adequately, we proceeded to evaluate the specific path coefficients.

Figure 2 shows that greater pornography use among women was not associated with any aspect of sexual satisfaction. Greater provocative, sexual media use was significantly associated with lower satisfaction with the love and affection in a sexual relationship (β = −.15, p = .01), lower satisfaction with the overall sexual relationship (β = −.14, p = .02), lower satisfaction with the variety in the sexual relationship (β = −.24, p < .01), and lower satisfaction with time spent on intercourse (β = −.14, p = .03).

OPEN IN VIEWER

Figure 2.

Structural equation model for women’s use of sexual content and sexual satisfaction. χ²(274) = 387.94, p < .01, CFI = .98, RMSEA = .03, SRMR = .03. Analyses controlled for age, race, sexual orientation, education, impulsivity, body image, religiosity, whether the participant had a child, and U.S. residency. Space constraints required abbreviations of several terms. “Foreplay” indicates satisfaction with time spent on foreplay. “Love” indicates satisfaction with love and affection. “Intercourse” indicates satisfaction with time spent on intercourse. *p < .05.

For control variables, higher impulsivity was associated with lower satisfaction with time spent on foreplay (β = −20, p = .01), variety (β = −.16, p < .01), frequency (β = −.18, p < .01), and time spent on intercourse (β = −.21, p < .01). Higher body image was associated with higher satisfaction with love and affection (β = .26, p < .01), frequency (β = .21, p < .01), overall (β = .26, p < .01), variety (β = .15, p = .03), and time spent on foreplay (β = .15, p = .01). Non-White women reported lower satisfaction with time spent on foreplay (β = −.13, p = .02), overall (β = −.18, p < .01), and frequency (β = −.18, p < .01). Those who were not residents of the U.S. reported higher satisfaction with variety (β = .20, p < .01) and frequency (β = .12, p = .04). Higher religiosity was associated with higher satisfaction with time spent on foreplay (β = .14, p < .01) and time spent on intercourse (β = .11, p = .03). Those who had a child reported less satisfaction with love and affection (β = −.20, p < .01). Those in a nonheterosexual relationship reported higher satisfaction with time spent on foreplay (β = .10, p = .03), and those with a higher education reported higher satisfaction with love and affection (β = .12, p = .02). Age was not associated with any of the dependent variables when all variables were evaluated simultaneously.

All covariances between sexual satisfaction items were estimated in the final model and varied between .19, p < .05 and .55, p < .05 (full results available from the first author upon request.)

Limitations and future directions

Although our study had many strengths, such as advancing conceptualization of sexual content and sexual satisfaction, a large sample size, and novel use of measurement, the current study has several limitations to consider. One limitation was the use of cross-sectional data. Longitudinal research is needed to establish the directionality of the results and clarify the explanations behind these associations. In particular, clarification is needed for how much use of sexual content influences different aspects of sexual satisfaction and to what extent the sexual relationship leads to use of pornography or provocative sexual media.

We also had limitations in measurement, as we only used one item for each dimension of sexual satisfaction. One of the dangers of one-item measurement is that we have no way to fully account for the reliability or validity of the constructs (DeVellis, 1991). While our study highlights that some distinction may exist in different aspects of sexuality, the one-item measurement limits confidence in what is truly being measured. Future researchers should create expanded scales from the multiple dimensions of sexuality that Pascoal et al. (2014) found to test the validity of these results. One aspect of sexual satisfaction our measurement did not assess that would be valuable to consider in future research is pleasure (Abramson & Pinkerton, 2002). Despite this limitation, it is worth noting that one-item measurement has been successfully used in studies on health (DeSalvo et al., 2006), work satisfaction (Wanous & Hudy, 2001; Wanous, Reichers, & Hudy, 1997), and self-esteem (Robins, Hendin, & Trzesniewski, 2001. Also, one-item measurement in sexual satisfaction research has a surprising amount of validity in comparison to traditionally validated sexual satisfaction scales (Mark, Herbenick, Fortenberry, Sanders, & Reece, 2014).

The items for sexual content also had limitations. We did our best to create scales for pornography use and provocative sexual media use that captured internal reliability within each scale yet also could be distinguished from each other. Overall, we believe that we accomplished this purpose in a way that suited our research question. However, it is important to note that within both pornography and sexual media are a wide variety of depictions of sexuality that we could not fully account for with our measurement. Some research questions may be better suited to evaluating highly specific sexual content that may influence a particular behavior (e.g., depiction of anal sex influencing the likelihood of engaging in anal sex). Additionally, our sexual media items focused specifically on the provocative portrayal of women. Future research should explore how sexual media focused on men differentiates from the focus on women, especially across gender.

Although we were able to establish distinction between pornography use and provocative sexual media use through confirmatory factor analyses, we note that the constructs still did have a moderate (for men) to high (for women) correlation between the scales. This suggests that although distinction exists between the types of content being viewed, when someone is viewing one type of content they also tend to view the other. Both pornography use and sexual media use seem to be tapping into an overall construct of viewing sexual content in general. The possibility also exists that some differences may be found between the items within the scales we used. For example, one item for the pornography scale focused on a solitary sexual act (masturbation), while the other items focused on dyadic acts. Assessing the differences between portrayal of solitary and dyadic acts could be yet another profitable area of future research.

There are also a variety of contextual factors that future research should account for. A clearer lens of the context could be evaluated through the perspective of both members of the relationship, especially since partner discrepancies have been found to be uniquely associated with relational outcomes in pornography research (Willoughby, Carroll, Busby, & Brown, 2016). Future research should explore how one partner’s use of pornography and provocative sexual media influences his or her partner’s multiple dimensions of sexual satisfaction. Other contextual factors worthy of future evaluation include relationship length, sexual orientation, and culture.

Many of our control variables should also be considered in the future as potential mediators and moderators of associations. For example, perhaps body shaming acts as a mediator or moderator between women’s use of provocative sexual media and sexual satisfaction. Body shaming may be the explanation for why viewing provocative sexual media influences sexual satisfaction, or it may act as a moderator, where objectifying sexual content only influences sexual satisfaction insofar as it leads a woman to feel insecure about her body.

Implications and conclusion

This study should be of use to researchers, as our factor analyses and the predictive validity of some of the associations provide a major step forward in acknowledging the multidimensionality of sexual content and sexual satisfaction. While some research has previously suggested distinctions between paraphilic (e.g., violent/coercive) and mainstream (e.g., vaginal or oral sex) pornography (Hald & Stulhofer, 2016), our study highlights that mainstream pornography can be distinguished from provocative sexual media. Researchers now have three distinguishable types of sexual content that they can evaluate in their studies and should feel free to explore the possibilities of different unique aspects of sexual content (e.g., same-sex acts, group sex). Furthermore, our work highlights that several aspects of sexual satisfaction are distinguishable and that there is danger in assessing several aspects as one monolithic construct. Future research should carefully consider whether their measurement of “sexual satisfaction” truly fits as a global appraisal of the sexual relationship, or if the items on the scale are actually differentiable, and should be evaluated separately.

To our knowledge, we have provided the first ever assessment of measurement invariance for men and women viewing sexual content. The lack of measurement invariance supports previous meta-analytical evidence (Oliver & Hyde, 1993; Petersen & Hyde, 2010) suggesting substantial differences between men and women when it comes to solitary sexual behavior. Men and women perceive items asking about sexual content through a fundamentally different lens. For example, when they both see an item such as “A video showing two naked women or men manually stimulating each other,” they are not thinking about this item in the same way. This has implications for the state of the research on sexual content, as it calls into question whether gender difference findings for men and women’s use of sexual content are due to underlying measurement problems that need further exploration.

This study should also be of use to therapists, as they may find it useful to consider with more detail what sexual expectations couples hold for satisfaction in their relationship. Therapists and relationship educators may also find this encouraging to consider more fully what exact sexual content couples are viewing. Therapists and couples should have detailed consideration of how different types of content may be influencing their scripts surrounding sexuality and the extent to which those scripts are congruent or incongruent with both partners’ sexual desires. Also, when relationship educators inform couples about the potential influence of using sexual content in a relationship, they should be very clear about the importance of understanding what specific content is being viewed, and how differing types of content may influence different aspects of sexual satisfaction in different ways, based upon each individual’s expectations of sexuality in the relationship. For example, couples, researchers, therapists, and relationship educators may all consider the extent to which the sexual content presents a script of objectification, as objectification is one aspect of sexual content with clear empirically validated implications for individuals using the content (Zurbriggen et al., 2011).

Script theory and objectification theory provide strong frameworks that researchers can use in future studies and should provide direction to help therapists and relationship educators in efforts in helping couples to navigate use of sexual content in committed relationships. Naturally, the pioneering nature of our study resulted in several limitations. But it is a major step forward theoretically and empirically in opening up a wide range of possibilities for evaluating how different types of sexual content influence multiple aspects of sexual satisfaction.