Introduction

The objectification of women is pervasive in the United States (American Psychological Association 2007; Bartky 1990; Fredrickson and Roberts 1997) as well as other nations around the world (e.g., Crawford et al. 2009; Gill 2008; Lazar 2006). Empirical evidence for the objectification of women (mostly relying on samples from the United States) is centered in two areas. The first area of empirical evidence is the widespread phenomenon of the “male gaze,” wherein men direct prolonged, unreciprocated glances at women (Argyle and Williams 1969; Cary 1978; Fredrickson and Roberts 1997; Fromme and Beam 1974; Hall 1984), which are often accompanied by sexually evaluative comments (Allen 1984; Gardner 1980). The second area is the extensive sexualization of women’s bodies (or individual body parts) in the media, including the pervasive use of women in sexual poses, often to sell products (Gill 2008; Lazar 2006), and the literal separation of sexualized body parts from the rest of the female body (e.g., a feminine leg being used as the base of a lamp).

Objectification theory (Fredrickson and Roberts 1997; McKinley and Hyde 1996) is an influential feminist theory that describes the process whereby individuals who are subjected to such objectification come to internalize the perspective of the outsider, a phenomenon called “self-objectification.” Because objectification is often a gendered process (with women subject to the male gaze), self-objectification occurs more often in women than in men (e.g., Aubrey 2006; Fredrickson et al. 1998; McKinley 2006a), but can occur in men as well (Hebl et al. 2004). When self-objectification occurs, an individual focuses attention on how her body appears to others rather than on how her body feels and on how she can, using that body, perform actions in the world. The theory predicts several consequences of self-objectification, including body shame, anxiety, eating disorders, depression, and sexual dysfunction. Many of these predictions have been supported by subsequent empirical research, mostly on U.S. women (for a review, see Moradi and Huang 2008).

The objectification of other people is also implicated in a number of negative outcomes ranging from very severe (as when the objectification of the enemy leads to torture or atrocities during wartime; Moshman 2005; Zurbriggen 2008) to the more mundane (as when viewing objectified images of women causes men to feel less satisfied with their romantic partners; Zillmann and Bryant 1988). Anecdotal evidence suggests that men who hold objectifying beliefs about women may have difficulty forming intimate relationships with them (Brooks 1995) and at least one correlational study supports that conjecture, finding that men’s satisfaction with their romantic relationships is negatively associated with objectifying beliefs associated with traditional masculinity (Burn and Ward 2005). On the other hand, some theorists have suggested that romantic relationships are the one place where objectification is safe and perhaps even enjoyable (Nussbaum 1999).

Although intimate romantic relationships are clearly a rich and important site for studying the effects of self-objectification and the objectification of others, surprisingly little empirical research has focused on this domain. In addition to contributing to the theoretical understanding of objectification, an empirical focus on objectification in romantic relationships can highlight important consequences of a culture saturated with objectification. As well, a focus on romantic relationships leads naturally to an examination of objectification from both sides of the (gendered) coin—the self-objectification that many women experience and the objectification of female partners that is encouraged by male socialization. Our aims for this research project were to bring the study of objectification theory into the context of romantic relationships, to focus on both self-objectification and the objectification of one’s romantic partner, and to examine the role of consuming objectifying media images. We were especially interested in understanding the association between each of these factors and relationship satisfaction. To empirically investigate these research questions, we gathered self-report data from a sample of male and female undergraduate students in the United States. Because this project investigates objectification and relationships in a U.S. context, note that the literature review is purposefully focused on studies reporting data from U.S. samples; exceptions are noted.

Self-objectification

Self-objectification has been theorized to have many negative outcomes (Fredrickson and Roberts 1997; McKinley and Hyde 1996). For example, self-objectification was hypothesized to increase feelings of shame and anxiety about the body, to decrease awareness of internal bodily states, and to reduce the likelihood of being in the creative and pleasurable state of “flow” (Csikszentmihalyi 1990). These psychological states, in turn, were expected to be implicated in a variety of problems that women experience, including eating disorders, depression, and sexual dysfunction (Fredrickson and Roberts 1997); moreover, the gender difference in self-objectification was proposed as a key explanation for gender differences in these mental health problems. Subsequent empirical research has supported many of these predictions (for a review, see Moradi and Huang 2008). For example, women who self-objectify are more likely to show signs of eating pathology (Daubenmier 2005; Hurt et al. 2007; Moradi et al. 2005; Muehlenkamp and Saris-Baglama 2002; Noll and Fredrickson 1998; Tylka and Hill 2004), depression, (Grabe and Jackson 2009; Hurt et al. 2007; Muehlenkamp and Saris-Baglama 2002; Muehlenkamp et al. 2005), and lowered self-esteem (Breines et al. 2008; Hurt et al. 2007; Mercurio and Landry 2008). Experimental research suggests that self-objectification impairs women’s cognitive performance (Fredrickson et al. 1998; Quinn et al. 2006) and increases negative affect (Gapinski et al. 2003) and appearance anxiety (Roberts and Gettman 2004). Correlational research has demonstrated an association between self-objectification and body shame in both women and men (McKinley 2006a, b).

Although this wealth of research has solidified a connection between self-objectification and outcomes at the individual level, little research has been devoted to examining how self-objectification would operate in particular social contexts, such as within romantic relationships. This is surprising because objectification is inherently a social phenomenon, and self-objectification is understood to arise from the process of being objectified by other people.

The one social domain in which some research on self-objectification has occurred is in the area of sexuality. Fredrickson and Roberts (1997) made specific theoretically-grounded predictions about self-objectification and sexual dysfunction; namely, that self-objectification would lead to decreased sexual satisfaction. The hypothesized mediating process is that self-objectification leads to shame and anxiety, which in turn results in the inability to connect with internal bodily states, something that is centrally important for experiencing sexual pleasure. Results from several studies support this prediction. Roberts and Gettman (2004) experimentally induced a state of self-objectification in young men and women and found that, for women, this led to reduced interest in sexual relationships. In a correlational study, self-objectification was associated with lower levels of sexual assertiveness in 12th grade girls (Impett et al. 2006). In addition, several researchers have found links between sexual dysfunction and variables that are closely correlated with self-objectification, such as self-consciousness or body shame. Sanchez and Kiefer (2007) found that, in a sample of both men and women, the relationship between body shame and sexual problems was mediated by sexual self-consciousness during physical intimacy. Similarly, using an all-female Australian sample, Steer and Tiggemann (2008) found that self-consciousness during sex mediated the negative relationship between both body shame and appearance anxiety with sexual functioning. In the present study, we aim to add to this pattern of findings by testing for the presence of a negative relationship between self-objectification and sexual satisfaction, a variable that has not yet been investigated.

Although Fredrickson and Roberts (1997) theorized only about sexual behavior and sexual functioning rather than about romantic relationships more broadly, we believe that objectification theory can support predictions related to other aspects of romantic relationships. A focus on appearance and a third-person view of one’s own body is predicted to lead to anxiety and body shame. Because appearance and sexual attraction are generally accepted as being relevant to romantic relationships, the anxiety and shame from self-objectification may be heightened in the context of a romantic relationship, leading to decreased relationship satisfaction. Furthermore, a focus on one’s own appearance might interfere with time and attention devoted to forming a connection with a romantic partner. Finally, self-objectifying and thinking of oneself as an object that is to be admired and used by someone else might prohibit emotional connections with a partner because it does not allow room to assert one’s own needs and desires within the context of the relationship. In other words, self-objectification might include a focus on serving one’s partner (e.g., by maintaining an attractive physical appearance and being sexually available) to the exclusion of communicating one’s own relationship needs.

Empirical research on the association between self-objectification and relationship satisfaction has been sparse. Sanchez and Broccoli (2008) found a significant correlation between self-objectification and decreased relationship satisfaction in a sample of college women. They also found that priming romantic relationships led to an increase in self-objectification in single women, suggesting that there is a link between self-objectification and attempts to find a romantic partner. Downs et al. (2006) found a relationship between body shame (a variable closely related to self-objectification) and decreased relationship satisfaction in a sample of female exotic dancers, but not college women. In the present study, we examine the relationship between self-objectification and satisfaction in romantic relationships in a sample of both female and male U. S. college students, predicting that self-objectification will be associated with lower relationship satisfaction.

Partner-objectification

Much of the empirical research on objectification has focused on the consequences of self-objectification. However, objectification theory (Fredrickson and Roberts 1997) states that self-objectification is an internalization of the objectifying perspectives of other people; thus, objectification by others is hypothesized to precede self-objectification and is thus the more primary or foundational causal agent. Although a growing number of studies are attempting to articulate the processes whereby objectification by others is internalized as self-objectification (for a review, see Moradi and Huang 2008), much less research has focused on the direct consequences of objectifying other people. An important contribution of this paper is that it examines implications of objectification not only for the objectified, but also for those doing the objectification.

Because self-objectification is theorized to arise from the internalization of the habitual objectification of one’s body by others, it is likely that objectifying others is actually more pervasive than objectifying oneself. Indeed, Strelan and Hargreaves (2005), using a mixed-gender Australian sample, found that objectifying other people is actually a rather common experience. In fact, women are more likely to objectify other women than to objectify themselves. They also found that those who self-objectify are more likely to objectify others. Additionally, women are objectified more than men by both men and women. It seems likely, then, that the present study will show a relationship between self- and partner-objectification, such that the more individuals objectify themselves, the more they will objectify a partner. Furthermore, because women are objectified more than men, it is reasonable to predict that men’s rates of objectifying female partners will be higher than women’s partner-objectification of men and that women’s rates of self-objectification will be higher than men’s.

Objectifying others may have particular outcomes when the person objectified is a romantic partner. The emphasis on appearance and physical attraction in romantic relationships would seem to increase the probability that people will objectify their romantic partners. Sanchez et al. (2008) reported that, whereas women show more signs of body shame than men, men seem to express more concerns regarding their romantic partner’s appearance compared to women. Further, the more participants expressed concerns for their partner’s appearance, the less satisfied they were with their relationship. We theorize that this is due to thinking of one’s partner as an object, whose purpose is sexual pleasure, rather than as a thinking, feeling person. This objectification may preclude one from developing a more personal, emotional connection with one’s partner. Thus, in the present study, we hypothesize a similar pattern of results, wherein partner-objectification (operationalized by surveillance of partner’s appearance) will be negatively related to relationship satisfaction.

We will also test the relationship between partner-objectification and sexual satisfaction. Because physical appearance of one’s partner is one source of sexual desire, it is possible that sexual satisfaction is an area in which partner-objectification is actually beneficial. It may be the case that increased thinking about a partner’s physical appearance increases sexual desire and the quality of sexual experiences. On the other hand, as we hypothesized with relationship satisfaction, it is also possible that focusing on one’s partner’s physical appearance precludes consideration of his or her emotional (or sexual) needs. Furthermore, objectifying a romantic partner involves viewing one’s partner as an object for one’s own sexual desire, which may interfere with the intimacy often associated with sexual satisfaction (Brooks 1995). The present study will begin to tease apart these competing hypotheses by examining the relationship between partner-objectification and sexual satisfaction.

A Predictor of Self- and Partner-objectification: Consumption of Objectifying Media

Self- and partner-objectification may arise from a number of different sources. Fredrickson and Roberts (1997) identified the media as one important influence on self-objectification, because images from the media often focus on the body especially women’s bodies, in an objectifying manner. Subsequent research has found evidence for this connection (e.g., Aubrey 2006, 2007); for both women and men, increased exposure to objectifying media predicted increased self-objectification. Particular emphasis has been placed on the objectification of women’s bodies in magazines (e.g., Morry and Staska 2001). Some studies have failed to find a relationship between viewing television or listening to certain music and increased self-objectification, but have shown a positive relationship between reading magazines and self-objectification (e.g., Slater and Tiggemann 2006 [Australian sample]). Therefore, the present study investigates media consumption in general as well as by genre (e.g., television, music, and magazines).

When viewing media that objectify women, both men and women may internalize the message that women are sexual objects, whose worth should be based upon their appearance. Previous research has shown evidence for this process, both correlationally (Ferris et al. 2007; Gordon 2008; Peter and Valkenburg 2007 [Netherlands sample]; Zurbriggen and Morgan 2006) and experimentally (Kistler and Lee 2010; Ward and Friedman 2006; Ward et al. 2005). Consumption of objectifying media is theorized to contribute to self-objectification through an internalization of the media’s presentation of individuals as sexual objects (Fredrickson and Roberts 1997). This logic extends to partner-objectification, wherein viewing objectification may offer a lens for viewing one’s partner. In fact, because viewing objectifying media involves objectifying someone else (e.g., the model or actress), it may actually have a stronger association to partner-objectification, which also involves objectifying someone else, than self-objectification, which requires a leap to thinking about the self.

In the present study, we aim to replicate the relationship between consumption of objectifying media and self-objectification that has been found in previous studies (e.g., Slater and Tiggemann 2006). Furthermore, we hypothesize a similar relationship between consumption of objectifying media and partner-objectification, wherein increased media consumption is related to partner-objectification.

Finally, the inclusion of consumption of objectifying media as a variable in this study allows for a test of its association with relationship satisfaction. Specifically, we will test a path model in which consumption of objectifying media is related to (reduced) relationship satisfaction through self- and partner-objectification (see Fig. 1). This model includes two main predictions: 1) Consuming objectifying media will positively predict self-objectification and partner-objectification; 2) Self- and partner-objectification will be associated with lower levels of relationship satisfaction.

Fig. 1
figure 1

Path diagram showing hypothesized relationships predicting relationship satisfaction

Full size image

Summary of the Present Study

This study examines objectification as it relates to romantic relationships. Primary variables of interest include self-objectification, partner-objectification, consumption of objectifying media, relationship satisfaction, and sexual satisfaction. The study includes a few predictions that have been tested and confirmed in previous research, but also focuses on novel research questions, particularly concerning partner-objectification. Of note, this is one of the first studies to theorize and test a consequence of objectifying someone else (as opposed to oneself). Furthermore, the emphasis on romantic relationships is a rich and important context for studying objectification, due to the connections between appearance, sexuality, and romantic relationships. Finally, by including both men and women as participants, we can test for gender differences in the relationships among variables. Although there are no strong reasons to predict such differences, we test for this possibility in all analyses.

Using bivariate correlations, multiple regression analyses, and/or structural equation modeling, the following hypotheses will be tested in the present study.

  1. 1)

    Levels of partner-objectification will be higher in men than in women and levels of self-objectification will be higher in women than in men;

  2. 2)

    Self-objectification and partner-objectification will be positively correlated;

  3. 3)

    Self-objectification will be associated with lower levels of relationship and sexual satisfaction;

  4. 4)

    Partner-objectification will be associated with lower levels of relationship and sexual satisfaction;

  5. 5)

    Consumption of objectifying media will predict self- and partner-objectification;

  6. 6)

    A path model will connect consumption of objectifying media and relationship satisfaction through self- and partner-objectification.

Method

Participants

Participants were part of a larger two-wave longitudinal study on sexual socialization and sexual aggression conducted at a large public university on the west coast of the United States. They were assessed in the fall quarters of their first and second years of college; data analyzed for the present study were collected in the fall of participants’ second year. Participants were a subset of the 184 participants (79 men and 105 women) who participated in the second wave of the study. Twenty-three participants who had never been in a romantic relationship and two participants who did not complete the relationship satisfaction measure were excluded. This left a total of 91 women and 68 men who comprised the final sample reported in all analyses below.

By design, participants ranged in age from 18 to 20 years old (M = 19.04, SD = .34). Men (M = 19.13, SD = .38) were, on average, about 2 months older than women (M = 18.98, SD = .30). Although small in absolute terms, this difference was statistically reliable, t(157) = 2.86, p = .005. Therefore, all substantive analyses reported below were run with age as a covariate. In no case did age explain a significant amount of variance, nor did results differ from those produced from analyses in which age was not included. Accordingly, analyses presented below do not include age as a covariate. The majority of the sample (82.4%) identified as heterosexual, but eight participants identified as gay or lesbian and eight identified as bisexual. Twelve participants selected other, unsure, prefer not to identify, or did not respond in response to the sexual orientation question. A chi-square test found no significant gender difference in the frequency of participants reporting heterosexual vs. some other sexual orientation, χ2(1) = 1.81, p = .18.

The majority of the sample (67.9%) identified as White, followed by Asian/Pacific Islander/Indian (11.3%). There were also 13 individuals identifying as biracial/multiracial, 11 as Latino/Hispanic/Chicano, four as Middle Eastern, and three as African American. One woman selected “other” and one woman chose not to identify her ethnicity. A chi-square test found no statistically reliable gender difference in the frequency of participants reporting White vs. some other ethnicity, χ2(1) = 1.48, p = .22.

For most participants (n = 148; 93.1%), one or both parents had attended some college or had received a bachelor’s or graduate degree. Most of the participants were either single (n = 78) or had a steady partner (n = 54), but 19 reported they were dating, six were living with their partner, and two were engaged. None of the participants reported being married. There were no statistically reliable gender differences in parental education or relationship status.

Measures/Instruments

Objectifying Media

To create a variable that reflects the participants’ consumption of objectifying media, we used a procedure similar to the one reported by Aubrey (2006). Participants rated how often they viewed various genres of television shows, films, magazines, and internet sites. Participants rated 16 television genres, ten film genres, eight magazine genres, and nine internet genres. Several examples for each of the genres were provided, to ensure that participants were similarly anchored. The scales used for each type of media (e.g., television, movies, etc.) were then converted to hours per week of exposure so that a combined measure could eventually be constructed.

Next, a list of these media genres was given to a mixed-gender group of ten experts (researchers working in a laboratory that focuses on objectification in the media). None of these experts were affiliated with the present project. The experts were asked to answer the following questions about each genre on a 5-point scale: 1) Within a prototypical episode[film, issue] in this genre, how frequently is it objectifying?, 2) When there is objectification in this genre, how intense is the objectification?, and 3) How familiar are you with this genre?. Each expert’s frequency and intensity ratings were multiplied to create one objectification rating for each genre from each expert. An overall weighted mean for each genre was computed by multiplying each expert’s objectification rating by his/her familiarity rating and dividing by the sum of the ten familiarity ratings (see Table 1 for weighted means and standard deviations for each genre). The final objectification scores thus had a possible range from 1–25.

Table 1 Levels of objectification in genres of television, film, and magazines
Full size table

The participants’ reports of how many hours per week they consumed each genre in each type of media were multiplied by the weighted mean of the objectification ratings from the experts. This resulted in each participant having a score for each genre that took into account amount of exposure and the frequency and intensity of the objectification present in that genre. Finally, these scores were summed to create one final score that represents the amount of objectifying media each participant consumed on a weekly basis.

Self- and Partner-objectification

Self-objectification and partner-objectification were assessed using the surveillance subscale of McKinley and Hyde’s (1996) Objectified Body Consciousness Scale. To measure partner-objectification, items were reworded so that all references to “I,” “my body,” or “how I look” instead referred to “my partner,” “my partner’s body,” and “how my partner looks.” Responses were measured on a 1-to-7 Likert-type scale (disagree strongly/moderately/mildly, neither agree nor disagree, agree mildly/moderately/strongly). Partner-objectification items were presented together, as a set. Self-objectification items were intermixed with items about the self from other scales (not analyzed for the present report). The coefficient alphas for self-and partner-objectification were .84 and .67, respectively.

Relationship Satisfaction

Relationship satisfaction was measured using the Relationship Assessment Scale (Hendrick 1988; Hendrick et al. 1998). This unidimensional 7-item scale measures overall satisfaction in a romantic relationship and includes items such as “In general, how satisfied are you with your relationship?”, “How good is your relationship compared to most?”, and “How many problems are there in your relationship?”. We used the same instructions described in Burn and Ward (2005); participants responded based on their current or most recent romantic relationship or (if they had never been in a romantic relationship) based on their best friend of the other gender. For all analyses, we included data only from participants who responded concerning a romantic relationship. Responses were measured on a 1-to-5 Likert-type scale (several different anchors were used, depending on item content). Coefficient alpha in our sample was .89.

Sexual Satisfaction

Sexual satisfaction was measured with a single item: “How satisfied are you with the sexual part of your current relationship?” Responses were on a 1-to-5 Likert-type scale (very unsatisfied, unsatisfied, have no feelings either way, satisfied, very satisfied). Participants were instructed to answer this question only if they were currently in a “committed dating relationship.” A total of 90 participants (51 women and 39 men) provided data on this item.

Procedure

All participants were run individually by female research assistants who were juniors and seniors attending the same university as the participants. Participants first completed a reaction time computer task, took part in a semi-structured interview, and then completed a questionnaire which included all measures for the current study as well as additional measures not analyzed here. Total time for most participants was 2 to 3 hr; they were given refreshments and breaks. Participants completed the questionnaires in a small private room; the research assistant was not present during this phase of the procedure. All participants received $25 for participating in this second wave of the study, as well as a bonus ($0–10) based on their performance on the computer task. Participants who did not cancel their originally scheduled appointment received an additional $5 scheduling bonus.

Results

To begin analyzing the data, we examined the distributions of all variables. Because the media consumption variable had several severe outliers, we performed a log transformation on the variable. This resulted in a distribution that approximated the normal distribution. Substantive analyses reported below used this log transformed version of the media consumption variable rather than the raw scores.

The means and standard deviations for all variables for both male and female participants can be found in Table 2. Participants reported consuming an average of 15.98 hr of media per week (SD = 15.62). By dividing the mean score for the consumption of objectifying media variable (186.11) by the average hours of media consumption, one can see that the average objectification score for each hour of media consumed was 11.65 (on a scale from 1 to 25), reflecting moderate levels of objectification. For both men and women, the average scores for self-objectification and relationship satisfaction were above the midpoint of the respective scales, indicating mild to moderate levels of self-objectification and relationship satisfaction. The average score for partner-objectification was at, or slightly below, the midpoint of the scale, indicating neutrality (men) or mild non-endorsement (women) of partner-objectification.

Table 2 Descriptive statistics and gender differences
Full size table

Independent samples t-tests were conducted to test for gender differences on the measures; results are reported in Table 2. There were no differences between men and women regarding their consumption of objectifying media or self-objectification. However, men reported objectifying their partners more and women reported greater relationship satisfaction. These differences are statistically reliable, even after applying a Bonferonni correction for conducting multiple t-tests. Thus, our hypothesis concerning gender differences in objectification was partially supported.

Zero-order Correlations

Zero-order correlations between all measures were computed, separately for men and women. With one exception (discussed below), there were no significant gender differences in these correlations; thus, correlations for the combined sample of men and women are reported in Table 3. Gender was significantly correlated with partner-objectification and relationship satisfaction, as reflected in the t-tests described above. Consistent with our hypothesis, partner- and self-objectification were positively related. Also consistent with our hypothesis, partner-objectification was negatively correlated with relationship satisfaction. Relationship satisfaction was also negatively related to both consumption of objectifying media (marginally reliably) and self-objectification. Finally, consumption of objectifying media was positively related to partner-objectification, although it was surprisingly not significantly related to self-objectification.

Table 3 Zero-order correlations between variables
Full size table

Fisher’s r-to-z transformation was used to test for gender differences in each of these correlations. Only one statistically reliable difference was found. The size of the correlation between self-objectification and partner-objectification in men (r = .547, n = 68) was larger than it was for women (r = .185, n = 91), z = 2.61, p = .009.

Predicting Relationship Satisfaction

To further examine the relationships between consumption of objectifying media, self-objectification, partner-objectification, and relationship satisfaction, we ran regression analyses predicting relationship satisfaction from the other three variables. Analyses were conducted separately for men and women, but the results were very similar. To confirm that there were no interactions with gender, we ran an analysis including regression terms representing all possible interactions with gender. None were statistically reliable (all t’s <1.20, all p’s >.25). We also tested for interactions with relationship status. None were found (all t’s <1.25, all p’s >.22). Accordingly, data for men and women were collapsed in further analyses, as were data for those who are or are not currently in a relationship. Possible multicollinearity was assessed by computing and examining variance-inflation factor and condition index statistics; no problems with multicollinearity were found.

Gender was included as a covariate because it correlated with relationship satisfaction (see above). Relationship status was also included as a covariate, because participants currently in a relationship reported greater relationship satisfaction (M = 4.17, SD = .71) than those participants who were not currently in a romantic relationship and thus were instructed to reflect on a past relationship (M = 3.19, SD = .87); t (157) = −7.82, p < .001. The other predictors were self-objectification, partner-objectification, media consumption, and all interactions between these three predictors. Results are presented in Table 4. Overall, the model predicted 36.0% of the variance in relationship satisfaction, F(9, 149) = 9.33, p < .0001. Partner-objectification was a significant predictor of relationship satisfaction, even controlling for the other variables in the model; higher levels of partner-objectification were associated with lower levels of relationship satisfaction.

Table 4 Unstandardized partial coefficients (B’s), standard errors, and standardized partial coefficients (β’s) for simultaneous regression predicting relationship satisfaction (n = 159)
Full size table

In order to test the proposed path model (Fig. 1), we conducted structural equation modeling analyses using EQS 6.1 with maximum likelihood estimation. Because there were no interactions with gender in the regression analysis, data for men and women were combined. To begin, a saturated model was estimated. In this analysis, gender was allowed to correlate with log consumption of objectifying media; both were predictors for all three endogenous variables. In addition, because of the high probability of shared method variance from similar item formats, the error terms for self-objectification and partner-objectification were allowed to correlate. Because this model was saturated, fit was perfect. This analysis confirmed the need to control for gender when predicting objectification of partner, as well as the absence of a gender difference in self-objectification. However, in contrast to the t-test results reported above, there was no significant path from gender to relationship satisfaction. We therefore dropped the (non-significant) paths from gender to self-objectification and from gender to relationship satisfaction to produce our final model (Fig. 2). Fit for the model was excellent; χ2(2) = .96, p = .62; NFI = .99, CFI = 1.00, IFI = 1.01, MFI = 1.00, GFI = 1.00, standardized RMR = .02, RMSEA = .00 (CI = .00, .13). Predictor variables accounted for 22.7% of the variance in objectification of partner and 15.3% of the variance in relationship satisfaction (only a negligible amount of variance in self-objectification was accounted for by the model).

Fig. 2
figure 2

Final path model predicting relationship satisfaction. Note: Error terms for self- and partner-objectification were allowed to correlate in order to model shared method variance. Paths in bold have coefficients that differ from zero, p < .05. Solid paths have coefficients that differ from zero, p < .10. Dashed paths have coefficients that do not differ reliably from zero, p > .10. Fit statistics: χ2(2) = .96, p = .62; CFI = 1.00, NFI = .99, RMSEA = .00 (90% CI: .000, .127)

Full size image

Consumption of objectifying media was marginally associated with objectification of partner (t = 1.925, p = .06) and objectification of partner was associated with lower levels of relationship satisfaction (t = −4.44, p < .0001). This indirect path was marginally reliable, according to a Sobel test (z = 1.77, p = .08). Other paths leading to relationship satisfaction were not statistically reliable. In particular, there was no direct path from consumption of objectifying media to relationship satisfaction (t = −.937, p = .35) and no indirect path through self-objectification (Sobel z = −.53, p = .60).

Predicting Sexual Satisfaction

To test our hypothesis that increased objectification would be related to decreased sexual satisfaction, we ran several analyses. A subset of 90 participants (31 men and 59 women) provided data on this item (those that were in a committed dating relationship). We first calculated zero-order correlations, separately by gender, and used Fisher’s r-to-z transformation to test for gender differences in the size of these correlations. Men had a statistically reliable negative correlation between sexual satisfaction and self-objectification (r = −.520, n = 31, p = .003). Although the direction of association was the same for women, the correlation was not statistically reliable (r = −.110, n = 59, p = .405). The correlation for men was larger, z = 2.01, p = .044. Similarly, men had a statistically reliable negative correlation between sexual satisfaction and partner-objectification (r = −.440, n = 31, p = .013), but the correlation for women was not reliable (r = −.144, n = 59, p = .276). However, the gender difference was not reliable, z = 1.41, p = .16.

Because partner-objectification and self-objectification were correlated, we ran a simultaneous regression analysis (separately by gender) to assess the joint and partialled effects of these two variables. No problems with multicollinearity were encountered. The results of these analyses are presented in Table 5. For men, self- and partner-objectification jointly explained a significant amount of variance (29.7%) in sexual satisfaction. Self-objectification appeared to be the stronger predictor, because it remained (marginally) reliable even after controlling for partner-objectification. These results must be interpreted with caution because of the small sample size. For women, although the partial regression coefficients were in the same direction as for the men, they could not be reliably distinguished from a null effect. This suggests that, for women, self-objectification and partner-objectification are either not related to or are more weakly related to sexual satisfaction.

Table 5 Standardized partial coefficients (β’s) for simultaneous regressions predicting sexual satisfaction, separately by gender
Full size table

Additional Analyses: Self-objectification and Specific Media Genres

Most previous research on media consumption and self-objectification has focused on the consumption of magazines (particularly those, such as fashion magazines, that focus on the “thin ideal” for women; e.g., Morry and Staska 2001; Slater and Tiggemann 2006). To facilitate comparisons with those previous studies, we ran several analyses examining consumption of specific genres of media, rather than total media consumption. In these analyses, self-objectification was significantly correlated with consumption of objectifying magazines (n = 158, r = .196, p = .014), but not consumption of objectifying television (n = 158, r = .026, p = .750) or movies (n = 158, r = .009, p = .911). This replicates other findings (e.g., Slater and Tiggemann 2006) that found a link between objectification and magazines, but not necessarily other media types (e.g., television).

Discussion

This study provided an intriguing set of findings concerning objectification in the context of romantic relationships among young adult college students living in the United States. All of the hypotheses described above were fully or partially supported.

  1. 1)

    Men reported higher levels of partner-objectification than did women; however, women did not report higher levels of self-objectification than did men.

  2. 2)

    Self-objectification and partner-objectification were positively correlated; higher levels of partner-objectification were associated with higher levels of self-objectification. This relationship was stronger in men than it was in women.

  3. 3)

    Self-objectification was associated with lower levels of relationship and (for men) sexual satisfaction; however, these associations weakened or disappeared when controlling for partner-objectification.

  4. 4)

    Partner-objectification was associated with lower levels of relationship and (for men) sexual satisfaction, even after controlling for self-objectification.

  5. 5)

    Consumption of objectifying media was positively correlated with partner-objectification but not with self-objectification. However, consuming objectifying magazines was associated with increased self-objectification.

  6. 6)

    A path model marginally connected consumption of objectifying media and relationship satisfaction through partner-objectification; the overall mediational model was partially supported. The evidence suggested the presence of an indirect path from objectifying media consumption through partner-objectification to lowered levels of relationship satisfaction; there was no direct path from media consumption to relationship satisfaction. The path from objectifying media consumption to partner-objectification was small in size (and only marginally reliable), so this finding should be interpreted with caution (and replicated). However, our findings suggest that any effect of consuming objectifying media on satisfaction with one’s partner is mediated through attitudes about one’s partner (namely, objectifying him or her).

Self-objectification

It was surprising that there was no reliable gender difference in self-objectification. Although researchers have occasionally found no gender difference or even higher levels in men (Hebl et al. 2004), most studies reveal higher levels of self-objectification in women (Aubrey 2006; Fredrickson, et al. 1998; McKinley 2006a). Indeed, higher levels in women are clearly predicted by objectification theory, because of women’s greater exposure to objectification and the male gaze. On the other hand, some scholars have argued that objectification of men is increasing in that men feel increased pressures to be strong and muscular (Pope et al. 2000). To the extent that this is true, gender differences in self-objectification would be predicted to decrease in magnitude, and our finding might be an early harbinger of this sea change. Another possibility, however, is that the women in our sample had unusually low levels of self-objectification. Supporting this notion, the mean for women in our study was 4.65, slightly lower than the means for young adult U.S. women reported in other recent studies: 4.81 (Moradi et al. 2005), 5.02 (Mercurio and Landry 2008), and 4.96 (McKinley 2006b). Our study was conducted at a university known for its progressive liberal politics; perhaps our female participants scored unusually high on an unmeasured variable such as feminist identity that accounted for the lower scores on self-objectification.

Partner-objectification

Although some theorists might argue that focusing on how one’s romantic partner looks is acceptable in the confines of a romantic relationship (Nussbaum 1999), in our data partner-objectification was associated, for both men and women, with lowered relationship satisfaction. This finding suggests that viewing one’s partner as an object is not good for one’s relationship. Importantly, whereas previous research has focused mostly on the consequences of being objectified, this study represents a possible consequence of perpetrating objectification. Thus, objectification appears to have negative consequences for both the target and the perpetrator.

Although this study provides evidence that partner-objectification has negative consequences for relationships, it remains unclear exactly what the mechanism is that relates partner-objectification to relationship satisfaction. It may be that partner-objectification entails focusing on one’s partner’s physical appearance and sexual availability, which could preclude considering their needs and desires, leading to damage to the relationship. It could also be that objectifying one’s partner leads him or her to self-objectify, which has been shown to have many negative mental health consequences (e.g., Hurt et al. 2007) that could put strain on the relationship. Future research should focus on teasing apart these explanations to more fully explain the intriguing relationship between partner-objectification and relationship satisfaction.

Also, because the present study focused on young adults, it will be important to investigate the relationship between partner-objectification and relationship satisfaction among couples of various ages who have been involved in relationships of varying length. Furthermore, the present study did not assess how attractive the participants felt their partners were; it may be that people who are dissatisfied with their partner’s looks are the ones who think about them most, and that dissatisfaction with their partner’s appearance is related to generalized dissatisfaction with the relationship. Future research could take into account perceived attractiveness.

Implications for Self-objectification Theory

Self-objectification was negatively associated with relationship satisfaction in the zero-order correlations, but this relationship disappeared in the regression and path models. In other words, once partner-objectification was controlled, the negative association between self-objectification and relationship satisfaction disappeared. Very few studies on self-objectification have also looked at partner-objectification. Moreover, objectification theory does not explicitly consider the possibility that being exposed to objectification (as a recipient or bystander) might lead not only to self-objectification but also to an internalized belief that all people (not just oneself) should be judged by their appearance and treated as an object. Indeed, the positive correlation between self-objectification and partner-objectification is consistent with this causal theory: self-objectification and partner-objectification could be correlated because of a common cause (exposure to objectification, either in the media or through interpersonal interactions). Our findings suggest that it is essential to develop theory to explain the relationship between self-objectification and the objectification of other people. Moreover, as research on self-objectification moves into the arena of social relationships and processes, it will be important to control for partner-objectification and the objectification of other people such as peers.

The central feature of self-objectification is a focus on appearance (how I look) instead of capabilities or internal states (who I am, how I feel, what I can do). Given the known negative effects of “spectatoring” on sexual satisfaction (Barlow 1986), it directly follows that being in a self-objectifying state during sex would be expected to result in lower levels of sexual satisfaction. Our findings tentatively support an association between self-objectification and reduced sexual satisfaction (at least in men); however, our regression results suggest that this association might be partially mediated through partner-objectification rather than through the process described above. Future research to replicate this finding will be helpful in understanding the roles that self- and partner-objectification play in sexual encounters and the ways in which they can lead to reduced sexual satisfaction. It will also be important to further examine the role of gender in this relationship. In our data, there was no statistically reliable association in women between self- or partner-objectification and sexual satisfaction. Perhaps sexual objectification is so normalized for women in this context that its effects are weak at best. Or perhaps there are important moderator variables (such as feminist identity) that operate for women, but not for men. Our sample was small, so it is also possible that a relationship in women exists and we did not have enough statistical power to detect it. Future research can explore all of these possibilities.

Future researchers may also wish to compare sexual satisfaction among different levels of relationship commitment. One possibility hinges on the difference between sex in the context of a loving, supportive relationship in comparison to “hook-ups” or other casual sexual encounters. The sexual satisfaction analyses presented in this study included only participants who were currently in a committed relationship and were reporting on their sexual satisfaction in that relationship. It seems plausible that being in a relationship provides reassurance that one’s body and appearance are, not just acceptable, but actually highly desirable to one’s partner. Thus, self-objectifying thoughts might be unlikely to arise during sex with a committed partner—the relationship provides enough safety that the individuals can relax and enjoy what they are feeling, rather than worrying about how they look. Casual sex with a new partner seems less likely to provide that safety. Future research that compares those two types of sexual encounters would be helpful for understanding the relationship between sexual satisfaction and self-objectification.

Implications for the Role of the Media

There was no relationship between overall objectifying media consumption and self-objectification. However, we did find (replicating previous research) that consumption of objectifying magazines was associated with self-objectification. The special relationship between magazines and self-objectification could be due to several factors. For example, it may be that the type of objectification in magazines is more body-focused than is the objectification in other media genres. Many objectifying magazines also include very direct and explicit advice on how to change one’s looks to fit the societal ideal, which could very easily be interpreted as instructions to self-objectify. Comparatively, although they still may feature objectified images and idealized bodies, other genres (e.g., films, TV shows) are less directive and thus may be less likely to lead to self-objectification. Regardless, this finding might prove useful to researchers who are exploring the exact mechanisms that relate objectifying media and self-objectification.

We did find an association between objectifying media consumption and partner objectification. This finding is important because it highlights how viewing objectified images can have negative effects on the viewer; in this case, consuming objectifying media was associated with increased partner-objectification, which was related to decreased relationship satisfaction. Furthermore, this finding suggests that we could profitably continue research in this area. Unlike research on violence in the media and (to a lesser extent) research on sexually explicit content, research that focuses on objectification in the media is relatively rare. Although sexually explicit content is likely to also be objectifying, this need not be the case. Thus, research that focuses specifically on objectification is sorely needed. We invite researchers to take up the challenge of devising better ways to measure objectification in different genres of media and hope that the information in Table 1 will be a helpful start to this process.

The fact that objectifying media consumption was related to objectification of another person but not of the self is reminiscent of a variety of findings and theories in communication studies, including the third party effect (Davison 1983), the impersonal impact hypothesis (Tyler and Cook 1984), and the differential impact hypothesis (Snyder and Rouse 1995). Research in this area has demonstrated that people can separate their personal judgments from their judgments about other people or society as a whole, and that exposure to media (or some forms of media; Snyder and Rouse 1995) is more likely to affect judgments about others than judgments about the self. Although the phenomena these theories seek to explain do not map exactly onto our study, it may be that some variant was at work here, in that the consumption of objectifying media differentially affected attitudes about another person (partner objectification) as compared to attitudes about the self (self-objectification).

Limitations, Future Directions, and Conclusions

The participants in this study were drawn from a U.S. college sample. They were moderately diverse in some respects (e.g., ethnicity, relationship status) but not in others (e.g., age, parental education, nationality). Future research should examine objectification variables in other populations, including in people from other cultures and nationalities. Of special interest would be a study examining the role of self- and partner-objectification in long-standing committed relationships (e.g., marriage or cohabitation). The findings of Sanchez and Broccoli (2008) suggest that self-objectification may play a much more central role when one is seeking a relationship than when one is in a relationship. Although a sizeable number of participants in this study reported being in a committed relationship, these were mostly of relatively short duration, due to the young age of all participants. Understanding the (perhaps changing) role of objectification in romantic relationships that span many decades would be illuminating.

Additionally, exposure to objectifying media is obviously not the only source of self- and partner-objectification. Self-objectification may be particularly likely to develop through exposure to direct actions of objectification (e.g., catcalls or sexual harassment). Other variables, such as gender role ideology, sexist beliefs, or empathy, are also worthy of exploration in conjunction with partner-objectification. Note that the reliability for our measure of partner-objectification was slightly less than is generally desirable, and so these future investigations may consider developing other measures of partner-objectification.

There is a strong theoretical rationale underlying the path model (see Fig. 1); however, the data presented here are all correlational and so the strength of any arguments about causation is limited. Future research may explore the effect of self- and partner-objectification on relationship satisfaction in an experiment to better isolate any causal relationships.

This study emphasized the importance of examining both self- and partner-objectification in the study of romantic relationships. Our general finding was that objectification (of the self or of one’s partner) was associated with lower levels of relationship satisfaction and (for men) sexual satisfaction. Moreover, exposure to objectifying media was associated with higher levels of self- and partner-objectification. The implication is that objectification, in the media and elsewhere, is implicated in a broad array of negative effects and that internalizing these objectifying messages is likely to be harmful to intimate, romantic relationships.