Relationship Satisfaction in Young Adults: Gender and Love Dimensions

Men and women present differences that go beyond evolutionary mechanisms of reproduction and species maintenance; social and cultural dimensions are indicated as modeling agents of different configurations of romantic relationships. This study presents the results of relationship quality models based on Sternberg’s triangular love theory. There were 335 subjects involved in romantic relationships, of whom 190 (56.7%) were male and 145 (42.3%) were female. Mean age of the participants was 29 years (SD = 9.1 years). The results of the study point out that love components predict relationship satisfaction differently per gender. For women, the intimacy, passion and commitment variables are significant predictors, whereas for men the commitment variable was not significant.


Introduction
Do men and women think, act and feel the same way about love, sex, marriage and fidelity? The psychological triad of feeling, acting and thinking is a study object of the most diverse theories; within the research field of romantic relationships, to identify differences and similarities between men and women in vertexes of that triad makes it possible to understand sociocultural and evolutionary elements of the dynamics of romantic relationships in the human species.
The popular saying "men are from Mars and women are from Venus" enables to understand the popular imaginary relative to the differences between men and women in romantic relationships. According to Sprecher and Toro-Morn (2002) such differences tend to be exaggerated and misunderstood, but some studies found sex differences about attitudes and beliefs on love. Within the context of gender studies, multiple conceptions, scientific and political orientations are observed (Narvaz & Koller, 2006). In a general way the scientific perspectives on gender attribute psychosocial explanations about the behavior of men directed to attitudes linked to work, strength and individuality, whereas women are directed towards family and care (Spence & Helmreich, 1972;Whatley, 2008).
Specifically in the context of romantic relationships, studies based on Lee's attitudes model (Hendrick & Hendrick, 1986) signaled that men were more sexually permissive than women and had higher scores in the ludus love style (cognitive style of game-playing love) and women scored more in storge (cognitive style of friendship love) (De Andrade & Garcia, 2014;Hendrick & Hendrick, 1995). In face of historic, biological and psychosocial differences that men and women tackle, and stressing the changes that emerged with the increase of female workforce, new gender equity roles in many Western countries and new family configurations, this study aims at investigating the relationship between romantic relationship satisfaction and love components in Sternberg's (1986) model among Brazilian men and women.

Relationship Satisfaction
According to Mattson, Rogge, Johnson, Davidson, and Fincham (2013), relationship satisfaction is on the most important variables in romantic relationship research. Romantic relationship satisfaction corresponds to an individual's judgment about the positivity of his/her relationship (Arriaga, 2001;Avivi, Laurenceau, & Carver, 2009).
According to De Andrade and Garcia (2012), the construct can be accessed through one-dimensional models (general quality) but also through multidimensional perspectives (specific variables). Specific aspects of romantic relationships such as, for example, communication, commitment, love, intimacy and commitment can contribute with relative influence in assessment of the relationship quality.
As can be observed in the literature, relationship quality is a significant predictor of subjective well-being, happiness and life satisfaction (Diener & Lucas, 2000;Myers, 1992). Neto and Pinto (2015) demonstrate the positive relationship between satisfaction with love life (global assessment of the quality of love life) and life satisfaction (global assessment of quality of life) and affective well-being in a Portuguese sample. According to the study, people who think that they have satisfactory relationships commonly present emotions and affect that are more positive towards life, and such relationship is modified in older people. For the latter, inferior levels of the relationship are observed. People facing separation and break-up processes, and consequent low quality in their marital relationship, go through a process of suffering, pain and mourning (Féres-Carneiro, 2008). In other words, people satisfied with their romantic relationship perceived the other contexts of live positively.
Amongst the most relevant predictors of relationship satisfaction, good communication strategies can be mentioned (Overall, Fletcher, Simpson, & Sibley, 2009). Couples with effective communication benefit directly in terms of the maintenance of relationship quality, just as couples with ineffective strategies face a loss in quality and therefore have a perception of higher costs, less success in the relationship and more conflict (Christensen & Shenk, 1991). Within conflict-resolution strategies, Zacchilli, Hendrick, and Hendrick (2009)

signal that Eros and
Agape styles are associated with positive strategies for the resolution of conflicts. In contrast, due to low commitment and adventurous attitudes, Ludus is associated with destructive strategies such as domination and reactive interaction. Pragma and Storge, according to the same study, are love styles linked to commitment strategies.

Triangular Model of Love
Robert Sternberg's triangular love conception (1986,1989,1997,1998) is one of the most relevant theoretical models within the sphere of romantic relationships phenomena, according to Masuda's (2003) review;and Hatfield, Bensman, and Rapson (2012). The model proposes a structure with three essential components of the romantic relationship: intimacy, passion and decision/commitment. Those aspects are schematically organized and in their ensemble constitute a triangle that, according to their combination in terms of presence, absence and intensity of feelings, structure different possibilities of the expression of love (Sternberg, 1986;Yela, 2006). According to Aron and Westbay (1996), the hypothesis of the three components of love is present in various theories about the phenomenon. It is also associated with the perspective of the implicit theories on love.
In Sternberg's theory, intimacy is related to feelings of proximity, consideration, bonds, valuation of the love relationship and of the companion (Sternberg, 1986;Yela, 2006). Passion is related to the physical attraction aspect and sexual contact, including interaction aspects linked to the "expression of desires and needs" dyad (Hernandez & Oliveira, 2003, p. 60) and sex, involving behavioral, affective and cognitive registers of esteem. Finally, decision/commitment is the part of the love feeling that is responsible for the maintenance and decision of keeping oneself in the relationship, linked to the expression of support, love and consideration (Cassepp-Borges & Teodoro, 2007).
For example, the absence of one of the three basic components would correspond to a "lack of love". On the contrary, if all components are present there would be full and complete love (Sternberg, 1989). The presence of only intimacy would constitute an expression marked by intense affection, very close to a feeling of friendship.
The exclusive presence of passion would lead to intense sexual desire, which would be under risk of extinction with time. Love based only in decision/commitment would form an "empty love", composed by strong union, but little intimacy and physical attraction (Sternberg, 1986(Sternberg, , 1989. For Ainsworth and Baumeister (2012), passion operates as an emotion, whereas intimacy is a condition for the maintenance of the relationship. According to Kim and Hatfield (2004), couples that experience passionate love tend to also experience positive affects, whereas companionate relationships are more associated with life satisfaction. Baumeister and Bratslavsky (1999) signal that changes in intimacy and passion levels are mutually interdependent. As intimacy is a necessary component to the establishment of the relationship, with the passing of time it tends to increase, just as passion decreases and has its determined course of time (Berscheid, 1983). Lemieux and Hale (1999) demonstrate the importance of the passion, intimacy and commitment dimensions for the satisfaction of unmarried American men and women.
More recently, in Yela's (2006) study, Sternberg's model received a new configuration. Following that perspective, there would be a subdivision of the passion dimension, with the decomposition in erotic passion and romantic passion, corresponding to a new hypothesis: a tetra-factorial model of love.
Applications of the triangular model to the understanding of facets of romantic relationships are also seen in personality studies. In Engel, Olson, and Patrick (2002) the components of love were related to variables of the model of the big five personality factors. Results showed that the conscientiousness personality trait was a significant predictor of the passion, commitment and intimacy variables. According to the authors, that personality dimension characterizes aspects of self-control and responsibility, elements that in their ensemble favor partner involvement in a romantic relationship. In the same direction, Ahmetoglu, Swami, and Chamorro-Premuzic's (2010) study with a sample of English adults points out that the agreeableness trait is associated with the three love dimensions, whereas conscientiousness is related to commitment and intimacy.
Other than personality, age is also a discriminant variable of the triangular aspects of love. Results from Ahmetoglu, Swami, and Chamorro-Premuzic (2010) indicate that higher age is positively associated with commitment and intimacy, while passion is negatively related to it, i.e., younger people guide themselves more intensely by elements of attraction and sexuality and older ones favor aspects associated with security, commitment and depth of involvement.

Gender, Sex and Romantic Relationships
The components of the romantic dyad present differences between them in terms of the generation of quality and happiness in the relationship. Behavioral differences of men and women cannot be explained only by social and cultural origins (Berry, Poortinga, Segall, & Dasen, 1992). They are also a product of biological differences (Hatfield & Rapson, 2002;Sprecher & Toro-Morn, 2002).
Concerning a more social and cultural dimension of romantic relationships, studies demonstrate that people live differently the experience of love, going further than the biological sexual dimensions attached to the phenomenon (Fehr & Broughton, 2001;Sprecher & Toro-Morn, 2002). In general, women are more prone to think love associated with emotional commitment and security components, while men, on their turn, associate it with sexual commitment elements and the "pleasure of intercourse" (Buss, 2000;Cimbalo & Novell, 1993).
Based on an evolutionary psychology approach, Buss (1989) carried out a study with 37 cultures of 33 countries about the aspects that influence partner choice, which pointed out that in many cultures men prefer to get involved with younger women and the women, in contrast, seek association with people with higher age than them. Moreover, in the physical attraction aspect, the author showed that in most of the studied cultures, with the exception of India, Poland and Sweden, good appearance significantly predicts the choice of the woman partner, i.e. physical attraction is more strongly associated with the choice made by men. These tend to choose women with proportional circumferences between waist and hips, a sign of high fertility in Europe and in the United States (Zaadstra et al., 1993).
In terms of the beliefs about love, based on John Alan Lee's six attitudinal styles of love -Eros (erotic), Pragma (rational), Ludus (game play), Storge (friendship), Mania (passionate) and Agape (altruistic) (De Andrade & Garcia, 2014, Hendrick & Hendrick, 1986-studies in various cultures indicate that men tend to have a predominant outgoing and adventurous style in their romantic interactions (ludus), while women possess more rational (pragma), friendly and companion (storge), intense and uncontrolled (mania) profiles than men (Hendrick & Hendrick, 1995, 2000Sprecher & Toro-Morn, 2002). Fabes and Martin (1991) and Mirowsky and Ross (1995) complemented those findings on love schemes differences between sexes: men are perceived as less expressive and emotional, while women are seen as more sensitive, emotional and easily susceptible to emotions. In what concerns love components for men, these involve more passion, whereas for women companionship and intimacy are stressed, but they result from the commitment that is established in the relationship (Kim & Hatfield, 2004;Sternberg, 1997).
And how would be the relationship between the passion, intimacy and commitment components in men and women's satisfaction with the relationship? This study aims at predicting global romantic relationship satisfaction of men and women from variables linked to Sternberg's love components model (intimacy, passion and commitment).
To understand peculiarities and generalizations about romantic relationships in Brazilian young adults is something that is articulated with widely studied aspects in the field of relationships worldwide, and still little explored in that

Method Participants
In total, 335 subjects took part of the study. Among them, 190 (56.7%) were male and 145 (42.3%) were female.
A proportion of 70% of participants were university undergraduate students, and the others had concluded high school. Mean participant age was 29 years (SD = 9.1 years). A condition for inclusion in the study was the fact that the individual declared to be involved in a romantic relationship at the moment of data collection. The mean relationship duration was 77.2 months (SD = 80.2 months).

Instruments
The instrument employed in data collection had demographic questions for participant characterization (sex, age, relationship duration, state of residence, and university course or occupation), as well as two psychometric instruments that had already been validated in a Brazilian context: 1) Reduced version of the Triangular Love Scale in Brazilian Portuguese (ETAS) validated by Cassepp-Borges and Teodoro (2007): this measure has 3 subscales that assess the components of love according to Sternberg's theoretical model (1986), intimacy (proximity and bond), commitment (decision to be inside the relationship) and passion (physical and sexual attraction). In that instrument, the participants judges his/her perception on the degree of intensity of the love aspects in the assessed relationship (in the current research, the romantic relationship existing at the time of data collection). Such evaluation is given through filling in 18 items. which have Likert-scale format ranging from 1 = "strongly disagree" to 5 = "strongly agree. Scores close to 5 indicate a high quantity, for the individual, of the pertinent trait in the relationship.
Cronbach's alpha coefficients equal or higher than .90. 2) Brazilian version of a general scale of relationship satisfaction (Schumm et al., 1986) and "I am satisfied with my relationship with my companion"] completed in a 5-point Likert scale ranging from 1 = "strongly disagree" to 5 = "strongly agree".

Data Collection Procedures
In conformity with the norms of the Brazilian National Health Council, the study was approved by the research with human beings ethics committee of the Federal University of Santa Catarina, under protocol number 84/06.
With the approval, data collection was conducted in public spaces of university campi and university classrooms of institutions located in the cities of Vitória and Porto Alegre, two Brazilian state capitals, by means of collective and individual questionnaire administration. Upon assessing that a participant was available for the data collection, i.e. was not busy with other activities, a researcher approached him/her and invited the person to participate. After agreement on the part of the subject, the questionnaire was handed in to him/her to be completed on the spot, with the availability of the researcher to clear any eventual doubts related to the activity.
Data were analyzed with the help of R software (R Development Core Team, 2010). The technique employed to characterize relationships between the involved variables was multiple regression.

Results
Prior to the main analyses, the reliability and validity of the employed measures were verified. The scales obtained the same factor structures observed in the original studies and the following Cronbach's alpha indicators with the Brazilian sample: Etas Intimacy: .84, Etas Passion: .84, Etas Commitment: .87 and Relationship satisfaction: .90.
At first a regression model with the general relationship satisfaction scale as the criterion and seven predictors: the Triangular Love sub-scales (Passion, Commitment, Intimacy) a dummy sex variable (men participants coded 0 as the reference group, and women coded as 1), and the interaction terms between the sex variable and the love sub-scales. In all regression analyses, multi-item measures were centered. Table 1 presents the descriptive data for the variables included in the model. Additionally, the correlation between any two variables did not reach the mean scale reliability, which shows no evidence of multicollinearity according to Campbell and Fiske (1959). The first model explained 48.83% of the variance of relationship satisfaction (F(7, 327) = 44.58, p < .001). Since the only interaction term that was significant was sex × commitment (t = 3.47, p < .001), a second model was run with the three ETAS sub-scales, the sex variable and only the sex × commitment interaction term. Almost the same amount of variance was explained with the second model (R 2 = 48.83, F(5, 329) = 62.73, p < .001).
Since the first model did not explain the dependent variable better than the second one (F(2, 329) = .08), p = .93), a decision was made to proceed with the second model, presented in Table 2. A third model was also tested by adding to the second model the participant age and relationship duration variables. Based on the same analytical procedures presented previously, the inclusion of such variables did not increase the prediction capacity of the model. The intimacy and passion dimensions were both significant positive predictors of relationship satisfaction. The interaction of sex and commitment was also significant; simple slopes analyses (Aiken & West, 1991) indicated that commitment was a significant positive predictor of relationship satisfaction for women participants (β = .43, p < .001), but not for men (β = .07, p = .22) (see Figure 1).

Discussion
Concerning the Brazilian version of the Triangular Love and Global Relationship Satisfaction measures, we observed good indicators for construct validity in this sample. Scale reliability indexes were good (all indicators are high than .84).
As for the hypotheses, H1 is supported by the results. Men and women from our Brazilian sample have complex differences and similarities in terms of the emotions and feelings towards a relationship. For the male sample the variables intimacy and passion indicate significant prediction of global satisfaction, whereas for women the three variables were significant predictors of perceived satisfaction.
As such, Hypothesis H2 is rejected: even though passion is a predictor, intimacy also contributed to the model.
It is thus observed that for male participants not only variables associated with physical and sexual attraction are important, but also that romantic interaction aspects associated with support, confidence and trust predict positive experiences in the relationship. As regards H3, it is rejected. For the sample of women, the satisfaction with the romantic relationship is associated with the concept of full love within Sternberg's original model (1986). The predictors of the global evaluation of relationship quality are the three dimensions of the love triangle: intimacy, passion and commitment. In the investigated sample, commitment had the highest means.
In the same direction, Stephenson, Ahrold, and Meston (2011), in a study about the reasons for sexual involvement and satisfaction point out that for women the array of variables in the satisfaction assessment is larger than for men. In addition, variables such as the commitment/love are significant predictors for both women and men, but with higher predictive value for the former.
Regarding the non-prediction of the commitment variable in the Brazilian men sample for general relationship satisfaction, such indicator does not signal its irrelevance. Such aspect apparently operates as a rule for men involved in stable relationships -dating and marriage-and is not a variable that is present in the assessment of relationship happiness. However, further studies are suggested to investigate the peculiarities of commitment in men involved in romantic relationships in Brazil, mainly its relationship with gender rules and stereotypes of men and women.
In a social psychological view, those differences might be associated with differences of gender and power in relationships. Kephart's (1967) study with a sample of Western students of different nationalities demonstrated how differences and similarities between men and women relate to the importance of what love is between the sexes.
In that study, over 1000 students were asked whether they would marry a person of the opposite sex if that person had all the expected qualities of a companion but one: they would not be in love with such person. The results pointed out that 65% of men would not accept marrying someone without love, while only 24% objected to such marriage.
Those results might be associated with sociocultural forces and patterns linked to attachments in Western society from that time period. Within the present study we did not ask subjects if they would marry with or without love, but rather we measured the intensity of love components that was present in the relationships of participants at the time of data collection. According to a study by Ogletree (2010)  from Texas, men and women have different attitudes toward long term relationships. Men are commonly guided by more traditions patterns of relationship configuration, whereas women focus more on planning. Gender differences are also signaled by Baber and Tucker (2006): men indicate less equality support in a relationship than women. However, in this study conducted with our Brazilian sample, the results contrasted with the ones presented by Ogletree (2010), in terms of participants' age and relationship duration. In the Brazilian sample, those variables did not increase the prediction of relationship quality when the properly affective aspects measured by ETAS were included in the model.
In Sternberg's (1986) model the concomitant presence of the passion, intimacy and commitment dimensions is given as a type of full love; the results of the study demonstrated a higher importance to women of a full love feeling, while for men the results are associated more with intimacy and passion, contrasting Kephart's (1967) results in a comparison. In the perspective of social structural theory (Wood & Eagly, 2002) the notions of different sexual interactions between men and women possess a relationship with the social roles of each sex in society.
In cultures with more equivalent sexual roles, as observed in some Western countries, the differences in terms of the constituting elements of the dyadic relation and the differences between men and women tends to occur with less frequency. That aspect was associated with the results of the present study in the intimacy dimension, important for both sexes, possibly due to a likely social and sexual equity among the participants of the study.
In general lines, as the historic review presented in Hatfield and Rapson (2005) emphasized, we observe various transformations in the context of romantic relationships in the last decades: less influence of the family in the matrimonial market, the growth of the women workforce, and the seeking for more professional satisfaction by couple members. Thus, as romantic relationships acquire a space of prediction of health and life satisfaction, future studies should investigate elements involving the difference in quality determinants in participant groups with different relationship durations. Sternberg (1998) pointed out that passion aspects have different configurations in short duration relationships; in such cases the evaluations are usually more elevated than in long term involvement.
Another variable that is clearly influential in the results of the found model is participants' culture and country of origin. According to Schmitt et al. (2009), Landis and O'Shea (2000) and Sternberg (1998), cultural dimensions are variables that mediate emotions and feelings.
The comparison of different ethnical and cultural groups is suggested for future studies, as well as the measurement of the referred constructs in a sample with couples with different relationship configurations and life stages. Likewise, inquiries employing multidimensional romantic relationship quality scales are suggested. Specific aspects of the love relationship, such as, for example communication and sex, contribute with relative influence in the global evaluation of the relationship (Fletcher, Simpson, & Thomas, 2000).
The choice of a one-dimensional measure of quality can be cited among the limitations of the study. It is known that such phenomenon has a multidimensional nature, and future studies shall investigate gender differences in those more specific models. Likewise, relationship duration, age and the existence of children might also moderate the role of gender in relationship satisfaction.
Additionally, passion demonstrated predictive value for both men and women. It is an emotion that is commonly associated with quality (Ratelle, Carbonneau, Vallerand, & Mageau, 2013) and suffers change through time (Ainsworth & Baumeister, 2012). However, perhaps due to a conservative component of women toward cultural and social components, that dimension might not be considered as important as are intimacy and commitment.