Compassionate Love for a Romantic Partner, Love Styles and Subjective Well-Being

Recently a compassionate love scale was developed to assess compassionate love or altruistic love for different targets (e.g., romantic partner, close others and all the humanity; Sprecher & Fehr, 2005). This study was conducted to examine the psychometric properties of the Compassionate Love Scale in the Portuguese context. In addition, it has been examined how compassionate love for a romantic partner was related to socio-demographic variables, love styles, and subjective well-being. Two hundred and eighty one men and women participated (42% of women) with a mean age of 21.89. All participants were currently in a romantic relationship. The Compassionate Love Scale shows satisfactory psychometric properties. Furthermore, our predictions were supported, as those who experience high levels of compassionate love for a romantic partner are more likely to report Eros and altruistic love (Agape), and subjective well-being.

peripheral ties, and all of humanity.The authors note that this definition is consistent with that of Lazarus (1991), who defined compassion as "being moved by another's suffering and wanting to help" (Lazarus, 1991, p. 289).Sprecher and Fehr (2005) used the term "compassionate love" rather than "compassion" in order to encompass emotional and transcendental nuisances, although they acknowledged that their construct could be named "altruistic love" or "compassion" (Sprecher & Fehr, 2005, p. 630).Although compassionate love may be experienced for someone to whom love is not reciprocated, it should not be confused with unrequited love, which is described in the literature as one-sided type of love (e.g., Baumeister, Wotman, & Stillwell, 1993).Sprecher and Fehr (2005) developed a scale to measure compassionate love for different targets (using alternative forms), ranging from humankind and strangers to relational partners.One advantage of this scale is that mean levels of compassionate love can be compared across targets.Another benefit is that predictors, correlates, and consequences can be compared across relational contexts.For example, the factors that may contribute to compassionate love for all of humanity may differ from those that contribute to compassionate love for a romantic partner.The Compassionate Love Scale has demonstrated high levels of internal consistency.For example, across three studies and across different versions of the scale, Cronbach's alpha has exceeded .90(Sprecher & Fehr, 2005; studies 1-3).In general, the scale is uncontaminated by social desirability biases.
Following an extensive review of the literature and an essentially qualitative analysis of intensive interviews with individuals across a wide age range and from different backgrounds, Lee (1973) proposed a comprehensive six-style model of love, with three primary styles and three secondary styles.The primary styles included Eros (passionate, romantic love), Ludus (game-playing love), and Storge (friendship-based love).Compounds of two of each of the primary styles formed the three secondary styles: Pragma (practical love, a compound of Storge and Ludus), Mania (possessive, dependent love, a compound of Eros and Ludus) and Agape (altruistic love, a compound of Eros and Storge).In Lee's model, an individual's love style is based on his or her preferences.As such, a person could have numerous preferences, each of which may be satisfied in a different relationship (Murthy & Rotzien, 1996).Lee wrote "it is the relationship that is styled, not the lover" (1988, p. 49).
Measures of love have been developed deductively and inductively grounded (Thompson & Borrello, 1992).The deductive inquiry approach is grounded in theory and is used in more traditional classical quantitative research.Other researchers take an inductive inquiry approach, which involves fewer specific hypotheses within the context of exploratory research, with an overall goal of developing new theory.The Hendrick and Hendrick (1986) measure was deductively grounded in Lee's typology reflecting only one approach.
The Hendrick and Hendrick Love Attitudes Scale was subjected to a crosscultural test by Neto (1993aNeto ( , 1994)).The love styles previously identified among U.S. students were also identified among students in Portugal.Although Lee suggested that men and women experienced love in numerous and different ways, the question still remains as to whether some styles of love are related to greater compassionate love than are others.
Subjective well-being focuses on how people evaluate their lives.Most researchers agree that there are three specific components of subjective well-being that can be operationalized and measured.Two of these components are within the affective domain and the other in the cognitive domain (Diener, 2000;Pavot & Diener, 2008).
The affective components include the presence of positive affect such as happiness or good feelings and the absence of negative affect.Negative affect refers to unpleasant feelings such as anxiety and anger (Argyle, 2001).Previous research revealed that love is an important predictor of happiness, satisfaction and positive affect (e.g., Diener & Lucas, 2000;Myers, 1992) Sprecher and Fehr (2006) showed that people reported experiencing positive benefits to the self as a result of experiencing compassionate love for others.
Participants perceived that their self-esteem, positive moral, self-awareness and spirituality increased as a result of feeling compassionate love for others.Thus this research pointed out that these are real benefits to the self of having an altruistic or compassionate orientation toward others.The design of this research focused on recalled data.The retrospective nature of this design "may result in participants being biased in recalling only those compassion love experiences that had a positive effect on the self" (Sprecher & Feher, 2006, p. 238).In the current research we will gather data on compassion love in daily life to determine whether such experiences are associated to subjective well-being.

Objectives of the present study
The primary aim of this study was to investigate the psychometric properties of the Compassionate Love Scale (CLS) for a romantic partner.An additional purpose was to examine how compassionate love for a romantic partner is related to sociodemographic variables, love styles and subjective well-being.Thus, the aims of the present study were three-fold.
Objective 1 -The first aim was to examine the psychometric properties of scores from the CLS (Sprecher & Fehr, 2005).The internal structure of the instrument would be scrutinized, and this scrutiny would include item analyses and exploratory factor analysis.
Objective 2 -The second aim was to examine whether there were differences in the compassionate love according to certain background characteristics, such as gender, religious involvement, time in relationship, and being in love now.Past research has documented gender differences in regards to compassionate love (Sprecher & Fehr, 2005), women reported experiencing compassion love for others to a greater degree than men, regardless of the target of compassion love.Marks and Song (2009) showed evidence across the compassionate norms and acts considered that women reported higher levels than men.
The Portuguese social environment is dominated by the Catholic tradition (Roman Catholic 96%; Protestant 1%, and Other 3%).Previous research has shown that religiosity was associated with agapic love style (Hendrick & Hendrick, 1992).Furthermore, spirituality and religiosity were associated positively with experiencing compassionate love (Sprecher & Fehr, 2005).At our knowledge no previous study has examined the effect of time in relationships on compassionate love.Thus we are not in position to advance a hypothesis about the effect of relationship on compassionate love.
"Falling in love with someone can reflect hormonal flux and physical attraction that can actually lead to giving of self for the good of the other" (Underwood, 2009, p. 5).We therefore predict that participants being in love experience more compassionate love.
Objective 3 -The third aim was to explore the relationship between scores on the Compassionate Love Scale with those on other relational constructs, such as love styles, and subjective well-being (satisfaction with life, love satisfaction, sex satisfaction, positive affect, and negative affect).
It was expected some love styles would be more strongly related to compassion love than would be others.Specifically, on the basis of the existing literature (e. g., Hendrick & Hendrick, 1992) several probable findings were anticipated.Eros (passionate, disclosing, high in self-esteem) is likely to want "an intense, exclusive focus on the partner but not possessiveness or jealousy" (Hendrick & Hendrick , 1992, p. 65).Ludus (game-playing, non-disclosing, sexually cynical) is love "without any particular serious intent" (Hendrick & Hendrick, 1992, p. 65).This is a manipulative, game-playing approach to love.Storge (friendship) seeks "a companionable, secure, trusting relationship with a partner who is similar in terms of attitudes and values" (Hendrick & Hendrick, 1992, p. 65).Thus similarity is very important for storgic approach to love, which may not be important for compassionate love.The pragmatic lover (logical, 'shopping list') "is more likely to have conditions before developing a relationship" (Hendrick & Hendrick, 1992, p. 66).Mania (possessive, dependent, lower in self-esteem) in which "The lover is jealous, full of doubt about the partner's sincerity and commitment, subject of physical symptoms such as inability to eat and sleep, experiences acute excitement alternating with debilitating depression" (Hendrick & Hendrick, 1992, p. 66).This is an obsessive-dependent approach to love.Finally, Agape (altruistic, committed, sexually idealistic) is "selfless and giving, concerned about the partner's welfare" (Hendrick & Hendrick, 1992, p. 66).Each love style includes aspects somewhat differently, leading to the following working hypothesis: compassionate love will be associated positively with Eros, and Agape, negatively with Ludus, and Pragma, and no association with Storge and Mania.
The cognitive component of subjective well-being was conceptualized as the intellectual evaluation of one's life satisfaction either globally or with respect to specific life domains.In the current study we will take into account global satisfaction and two specific life domains, love satisfaction and sex satisfaction.As discussed by Diener et al. (1985), life satisfaction stems from a judgmental process by the individual.In this vein sex life satisfaction and love life satisfaction can be considered as referring to cognitive, judgmental processes.
Love life satisfaction is a conscious cognitive judgment of one's love life in which the criteria for judgment are up to the person (Neto, 2005).Love satisfaction is an important component of well-being for most individuals (Salvatore & Munoz Sastre, 2001).Previous research linked love satisfaction to happinesss, self-esteem, eros, agape, romantic acts, and decreased loneliness (Neto, 2005).
Satisfaction with sex life can be defined as a global evaluation by the person of his or her sex life (Neto, 2012).Previous research linked reported sexual satisfaction to healthy disposition, life satisfaction (Apt, Hubert, Pierce, & White, 1996), and relationship satisfaction (Holmberg, Blair, & Phillips, 2010), Hence, we can predict that life satisfaction, love satisfaction, sex satisfaction and positive affect will be associated positively with compassionate love, and negative affect will be associated negatively with compassionate love.
The mean number of months the partners had been dating was 29.31 (SD = 26.16; the range was 1 month to 145 months).

Measures
The questionnaire consisted of scales described below, along with selfassessment and demographic questions pertaining to age, sex, and other aspects of relationships.One question asked participants, "Are you in love now?" and two questions were asked to evaluate the religious involvement: "Do you believe in God?" and "Do you attend church every week (except when you are truly unable to do so)?"All scales were previously adapted for Portuguese population, except the Compassionate Love Scale.In designing the Portuguese version of the Compassionate Love Scale, the author followed the guidelines proposed in the literature on crosscultural methodology (Brislin, 2000): independent/blind/back-translation, educated translation, and small-scale pre-tests.

(a)
Love Attitudes Scale.The 42-item Love Attitudes Scale (Hendrick & Hendrick, 1986;Neto, 1994) which measures the six major love styles described by Lee (1973) was used.The higher score signifying a higher value of that love styles variable.
(b) Satisfaction With Life Scale.This scale consists of five items (Diener et al., 1985) such as: "The conditions of my life are excellent."The reliability and the validity of this scale have been previously demonstrated for a Portuguese population (Neto, 1993b(Neto, , 1995)).Cronbach standardized alpha for the current study was .90.
(c) Positive and negative affect were assessed through the Portuguese version of the Positive and Negative Affect Schedule (Watson et al., 1988).This is a measure of positive and negative affectivity that includes 22 emotion adjectives.The scale has been extensively used and was adapted to a Portuguese population (Simões, 1993).For the current study the Cronbach standardized alphas were .87 and .88,respectively.
(d) The 5-item Satisfaction with Love Life Scale was previously developed (Neto, 2005).The scale consists of five items, such as "I am satisfied with my love life."Higher scores represent greater love life satisfaction.The reliability and the validity of this scale have been previously demonstrated for a Portuguese population (Neto, 2005).For the current study the Cronbach standardized alpha was .93.
(e) The 5-item Satisfaction with Sex Life Scale was also previously developed (Neto, 2012).The scale consists of five items, such as "So far I have gotten the important things I want in sex life."Higher scores represent greater sex life satisfaction.
The reliability and the validity of this scale have been previously demonstrated for a Portuguese population (Neto, 2012).For the current study the Cronbach standardized alpha was .93.
(f) Compassionate love for partner (Sprecher & Fehr, 2005).The 21-item relationshipspecific version of the Compassionate Love Scale (CLS) includes items such as, "I spent a lot of time concerned about the well-being of ___ [the partner]." Each item was followed by a 7point response scale (1 = not at all true to 7 = very true).

Procedure
Participants were recruited and tested by a trained psychology student.
Undergraduate students completed the questionnaire during or immediately following class time.Consent from the participants was obtained before the questionnaires were administered.Each participant answered individually in a quiet room at the university.
The experimenter was present when the participants filled in the questionnaires.
Anonymity and confidentiality were assured.Approximately 30 minutes were required to complete the questionnaire.

Factor Structure, Reliability and Descriptive Statistics of the Compassionate Love Scale
Prior to examining the internal consistency reliability of the CLS scores, a factor analysis was conducted using principal components analysis with varimax rotation.The significance of Bartlett's test of sphericity, Χ 2 (190) = 3767.87,p < .001,and the size of Kaiser-Meyer-Olkin measure of sample adequacy (KMO = .94)revealed that the items of the CLS had adequate common variance for factor analysis (Tabachnick & Fidell, 2007).The number of factors to be extracted was determined by an inspection of the scree plot of eigenvalues.This inspection showed one primary factor that explained 48.73% of the variance and that had eigenvalues of 9.75.Table 1 gives the respective component loadings on the first unrotated factor.
Cronbach's alpha for scores on the instrument was satisfactory, with alpha = 0.94 (Table 1).As could be expected on the basis of the results of the factor analysis and internal consistency analysis, the corrected item total correlation for each CLS item was substantial, i.e. over +0.40, except for item 13 ( r= .20).Predictably also the mean interitem r (homogeneity) for CLS item set was acceptable, 0.45.Thus, these findings, in conjunction with the exploratory factor analysis, suggest unidimensionality of the scale.
The mean score on the CLS was 5.27, with a standard deviation of 1.03.On a seven-point scale from low to high, this score indicates that compassionate love among these college students is moderately high, and more than the neutral mid-point of 4.00.Note.The response for each of the items ranged from 1 (not at all true of me) to 7 (very true of me).
Having shown that scores from the CLS had desirable psychometric properties, the next step was to examine the relationship between scores on CLS with those on love styles, and well-being measures to which one might expect it to be related, and background factors.(Totals of participants in the following paragraphs do not equal 281 because some of the participants did not respond to all the questions requesting background information.)

Background Variables and Compassionate Love
A number of specific questions pertaining to the participants' background were included to assess the potential suppositions about how compassionate love functions.
The approach that was used treated each background variable as an independent variable, and using participants' total scores on compassionate love scale as a dependent variable.One-way analyses of variance were performed on the data.The means and F ratio for CLS are shown in Table 2 for several background variables.The F ratio for each one-way analysis is shown at the top of the relevant column of means.
Religious involvement.The degree of religious involvement factor was created by combining responses from the belief in God and attendance in church items, with the levels: non-believer, believer/non-attendee, and attendee.There were religious involvement differences in compassionate love, [F(2, 269) = 4.64, p < .05,η 2 =.034)].
However, the effect size for the differences was very small.Using formulae taking into account differences in cell frequencies, pairwise comparisons of means showed the believers/regular attendees (n = 45, M = 5.60, SD= .91)scored higher than the nonbelievers/non-attenders (n = 89, M = 5.04, SD = 1.12) on compassionate love.By contrast, there were no significant differences between the believers/non-attendees (n = 136, M = 5.29, SD = .99)and the two other groups on compassionate love.
Relationship length.Relationship length was measured in months since the beginning of the current relationship.There was not a significant effect of time in

Correlations with Other Variables
The CLS was expected to correlate with other variables in predictable ways (Table 3).An inspection of the correlations indicates that CLS scores correlate significantly with all love styles, except Storge and Mania.Significant positive correlations were found with Eros and Agape, and significant negative correlations were found with Ludus, and Pragma.
As expected, all three satisfaction scores were related to scores on CLS.
Compassionate love was positively associated with the expression of satisfaction with life, satisfaction with love life, and satisfaction with sex life.The higher the compassionate love felt, the more satisfaction with life, with love, and with sex.
Compassionate love was also positively associated with the expression of positive affect.The higher the compassionate love felt, the more frequent the experience of positive affect.Conversely, compassionate love was not significantly associated with negative affect.

Discussion
The current research extended the study on compassionate love to another cultural context.Young adults completed the CLS with respect to a specific close other: a dating partner.First, we investigated whether the factor structure of responses to the CLS is consistent in the Portuguese culture.We found the same structure as was evidenced in the United States.In line with the results of Sprecher and Fehr (2005), two other factors with eigenvalues slightly above 1.0 and explaining 11% and 6% of the variance were extracted.We agree with the two reasons to present a scale as measuring a single factor advanced by Sprecher and Fehr (2005).First, the scree test demonstrated a distinct break between the first factor and the others.Second, the items loading on the second and third factors correlate with other variables considered in the study (e.g., love styles and subjective well-being) similarly to items in the first factor.Separate factors of a scale are less conceptually meaningful when they correlate in similar ways to other variables (Briggs & Cheek, 1986).
The Portuguese version of the 21-item scale showed good psychometric properties, as indicated by Cronbach's alpha, inter-item correlations, and item-to-total correlations.Attempts to validate the Portuguese version of the scale were met with success.
According to Underwood (2009) there are individual characteristics that encourage the expression of compassionate love in people.We have examined individual differences in propensity to experience compassionate love, focusing on variables such as gender, religious involvement, relationship length, being in love, love styles, and subjective well-being.
Gender differences in compassionate love were found.These findings are in line with previous research (Sprecher & Fehr, 2005;Hwang, Plante, & Lackey, 2008) in that women tended to score higher on the scale than men.This is also consistent with research showing that women express more empathy and emotional support for others, constructs which Sprecher and Fehr (2005) found related to compassionate love, than men (e.g., Eagly & Crowley, 1986).
As Argyle (2000, p. 194) pointed out "It is a central part of Christianity and of other religions to care for the poor, the weak, the sick, and outcasts of society, to show love, charity and compassion to them".Religious involvement, as predicted, did show a positive influence on compassionate love.The believers/regular attendees scored higher on compassionate love than the nonbelievers/non-attenders.The results on religious involvement differences tend to be consistent with previous research (Sprecher & Fehr, 2005).
In contrast, the relationship length had no impact on compassionate love.
Regardless of the duration of relationship, respondents showed similar compassionate love.We found that being in love does indeed change one's perspective.Participants in love displayed higher compassionate love than participants not in love.Lovers really do wear rose-colored glasses (Hendrick & Hendrick, 1988).Whether someone is in love or not appears to affect compassionate love.
The compassionate love showed significant relationship with many of the variables employed in the research and were for the most part consistent with predictions.Concerning love styles Eros and Agape were significantly and positively associated with CL, and Ludus and Pragma were significantly and negatively associated with CL.An additional set of hypotheses, that certain measures of well-being would be related to compassionate love, was globally confirmed.Indeed, compassionate love was positively associated with Eros.Eros represents the emotionally intense individual who is looking for a psychologically intimate and open relationship as well as a passionately expressive one.The Eros lover is self-affective for producing desired responses in others and tends to prefer to establish relatively close and exclusive romantic relationships.
Conceptually, the similarities between compassionate love and Agape rise the question of whether there are distinct constructs.Our findings indicate that although measures of these constructs are correlated, the correlations are not so substantial as to suggest that they are redundant.Based on the current findings it would appear that compassionate love is related to, but distinct from, the love style Agape.Lee (1988) described the typical game-playing as "often frustrated in adult life.
They are unwilling to commit themselves to love" (p.50).Thus, we hypothesized that scores on the CLS would be negatively correlated with a game-paying approach to love.
The results supported this prediction.Similarly, compassionate love was negatively associated with a "shopping-list" approach to love (pragma) in which relationships are formed with those who possess certain qualities (e.g., financial prospects, family background).Scores on the compassion love were unrelated to Mania and to Storge.
The more compassionate love participants experienced with their current relationship the more overall satisfaction.These findings are in agreement with studies showing an association between overall satisfaction and love (Neto, 2005).Current findings supported also this picture for two intimate domains satisfactions, as love satisfaction and sex satisfaction correlated with compassionate love.
Positive affect also appeared to be related to compassionate love.Empirical evidence indicated that the experience of subjective well-being can be beneficial to effective functioning.For example, individuals reporting high subjective well-being have stronger social relationships (Diener & Seligman, 2002).
There are several limitations and issues stemming from this study.First, although the CLS has been shown to have good reliability and validity in the current Portuguese version, it is a self-report instrument and therefore might be susceptible to social desirability bias.Second, the sample was drawn from undergraduate college students and therefore was not necessarily representative of the larger world of more mature people.Asking questions about compassion love of persons of different ages is essential to a fuller understanding of these issues.

Table 1 .
Means, standard deviations, factor loadings, and item-remainder correlations for the 21-Item Compassionate Love Scale for intimate partner version

Table 2 .
Means and F ratio for compassionate love as a function of selected background variables For each variable means with no subscript in common differed at the 0.05 level, either by F test directly for a pair of means or by the Scheffe test for three means.

Table 3 .
Correlations Between Compassionate Love Scale Scores and Other Variables