The univariate normality test revealed that the distribution of Relationship Assessment Scale (RAS) scores is not normal. However, asymmetry and kurtosis values were -1.029 and 0.115, respectively, which is not an extreme violation of normality (Finney & DiStefano, 2006). Correlation values between independent variables, regression coefficients and standard errors, respectively, were not high and, therefore, did not suggest the existence of multicollinearity. In addition, the Variance Inflation Factor (VIF) calculation showed no indicator values of multicollinearity, VIF > 5. In this study, the effect size was represented by the Coefficient of Determination (R²) and all interpretations of the effect sizes calculated in the MLRA followed the Cohen classification (1988).

For male G1 participants, the MLRA estimated that for each Standard Deviation (SD) of the predictor variable TLS-R Passion, there will be a 0.61 SD variation in the RAS Marital Satisfaction. In this case, the effect size of the explanatory variable on the criterion is strong (R² = 0.38). For female G1 participants, it was estimated that for each SD of the predictive variables Decision/Commitment of the TLS-R and the RBS Agape, respectively, there will be a variation of 0.62 and 0.38 SD in the Marital Satisfaction of the RAS (Table 1). The effect size of the two explanatory variables on the criterion variable was strong (R² = 0.57). For male G2 participants, it was estimated that the variation of a SD in the MSSI Proactive Self-Control and SSI Coping and Self-Assertion with Risk variables, respectively, will produce a variation of 0.77 and 0.47 SD in the RAS Marital Satisfaction. The effect size of these predictor variables was strong (R² = 0.68). It was also estimated that the variation of a SD in the EI Affective Sensitivity and the SSI Coping and Self-Assertion with Risk explanatory variables, respectively, will produce a variation of 0.58 and 0.32 SDs in the RAS Marital Satisfaction variable for women in the G2 group (Table 1). This was a strong size effect (R² = 0.35).

In the G3 group, it was estimated that the variation of a SD of the TLS-R Decision/Commitment explanatory variable will have an effect of 0.48 and 0.77 SDs, respectively, for men and women on the Marital Satisfaction criterion variable of the RAS. The sizes of these effects were medium (R² = 0.21) and strong (R² = 0.58), respectively, for men and women (Table 1).

The TLS-R Passion was the most prominent predictive variable for men in the G4 group. The variation of a SD in this variable should produce a variation of 0.88 SD in the Marital Satisfaction criterion variable. The effect size of this explanatory variable on the criterion variable was strong (R² = 0.76). For G4 women, it was estimated that the variation of one SD in the Passion variable of the TLS-R will have a respective effect of 0.71 SD on the Marital Satisfaction variable (Table 1). The effect size of this explanatory variable on the criterion variable was strong (R² = 0.50). In addition, the age of the participants had no significant effects identified in the overall Marital Satisfaction, nor in the four stages of the family life cycle, so it was excluded from the MLRA.

A two-way analysis of variance compared the average RAS scores between the family cycle stage variables and the participants’ sex. In the results, significant statistical differences for the main effect of the life cycle stage F (3, 440) = 6.784, p <0.001. Tukey’s follow-up test identified the differences between G1 (M = 4.44, SD = 0.59) – G2 (M = 4.42, SD = 0.47) and G1-G3 (M = 4, 12, SD = 0.62), and the effect sizes (ds) were moderate – 0.43 and 0.53, respectively (Cohen, 1988). Mean RAS scores were lower for the G3 group, of couples with adolescent children, followed by couples with young children. The highest scores were for couples without children and couples with adult children (Table 2). No significant differences were identified in the comparisons between the participants’ sexes and also in the interaction of sex with the family life cycle stages.

It was observed that the predictors of Marital Satisfaction for men and women varied from one stage of the family life cycle to another, which corroborates the idea that each of these moments has its own development, as described by McGoldrick et al. (2015). In the G1 and G2 groups, in which individuals mostly have a shorter marital relationship, the explanatory variables of Marital Satisfaction were different for men and women. On the other hand, the G3 and G4 groups, with longer-term relationships, had the same or similar predictors of Marital Satisfaction for both sexes (Table 1).

In the G1 group, men indicated the TLS-R Passion as the most important factor for Marital Satisfaction. This component of love is generally considered to lead to romance, physical attraction, sexual intercourse (Sternberg, 2006) and therefore plays an important role in the early times of romantic relationships. Few recent studies have reported sexual differences in Passion, with men showing higher levels than women (Ahmetoglu, Swami, & Chamorro-Premuzic, 2010) and when these differences appear, they are very subtle (Sumter, Valkenburg, & Peter, 2013).

For women in the G1 group, current data pointed to the TLS-R Decision Commitment and the RBS Agape predictor variables as responsible factors for Marital Satisfaction. The Decision/Commitment variable has a short-term dimension, a decision to love each other and a long-term commitment to maintain that love (Sternberg, 2006). These women indicated a tendency to be more satisfied with lovers who are predominantly committed and emotionally involved with the relationship, possibly for the relationship to be safer and more stable.

The relationship based on love, companionship, mutual admiration and individual freedom to build the life trajectory were the elements found in the maintenance of relationships by Borges, Magalhães, and Féres-Carneiro (2014). This style of love is considered as kind, loving and difficult to find, characterized by giving oneself to the partner. In the current data, it is possible to understand women’s responses as a search, at the beginning of relationships, for an affectionate, caring partner, for a friend, for commitment and companionship, which are characteristics of the TLS-R Commitment and the RBS Agape constructs. The main explanatory variables for male Marital Satisfaction in the G2 group were Proactive Self-Control and Coping and Self-Assertion with Risk. The Proactive Self-Control factor is characterized by the ability to keep track of emotions as well as reactions. This is essential for the good dialogue between the couple, as well as helping to recognize the emotional state of their spouses, listening to what they have to say, waiting for the right moment to express an opinion, and seeking for an understanding (Villa & Del Prette, 2012). Different studies have indicated that communication is important in romantic relationships (Lavner, Karney, & Bradbury, 2016; Schmidt, Luquet, & Gehlert, 2016). They suggest that when there is dialogue, there is an increase in satisfaction and intimacy. In the contemporary lifestyle, factors such as an excessive involvement with work and a lack of time can make communication and problem solving difficult (Alves-Silva, Scorsolini-Comin, & Santos, 2016).

For women in the G2 group, Affective Sensitivity appeared as the main predictor variable of Marital Satisfaction. According to Falcone et al. (2008), this factor is characterized by feelings of compassion and interest in the emotional state of the other person and, therefore, a higher score indicates a lot of attention or care towards the other. It should be noted that the G2 group is at the stage of the family cycle characterized by the arrival of children and the change in marital structure. Mothers tend to be more sensitive to hormonal changes, attention, and child-rearing concerns. This is a biological predisposition of women to protect and identify the needs of their babies. This could explain Affective Sensitivity as a predictor of Marital Satisfaction for women in this study.

Coping and Self-Assertion with Risk was the second factor that most affected Male and Female Marital Satisfaction in the G2 group. It portrays interpersonal episodes that lead the individual to react in defense of their own rights, the other people’s rights and self-esteem, and may have an undesirable reaction (A. Del Prette & Del Prette, 2001). In the study by Mosmann and Falcke (2011), couples claimed that the most frequent reason for fights, clashes, stress, conflicts, misunderstanding, and coping between them was raising their children, followed by time spent together, money, household chores, sex and legal issues.

The results of the G2 group demonstrate that, at this stage of the marital life, the couple may present differences in their perceptions. Husbands have a hard time understanding that they no longer have the exclusive attention from their wives, and women are overloaded with learning how to be both a mother and a wife. Men and women respond differently to changing family dynamics and new responsibilities to their spouse and children. Typically, these differences lead to conflict and dissatisfaction, and even divorce. Bäckström, Karing;reholtd, Thorstenssona, Golsäterc, and Maring;rtensson (2018) pointed the following factors as causes for the higher levels of marital conflict: the fact that women are more vulnerable to depression; women do most household and child care chores; men taking refuge in the workplace. or other activities that make them stay away from home; and a decreased frequency of dialogue and sex.

In the G3 group, men and women indicated the TLS-R Decision/Commitment predictor variable as the main predictor of Marital Satisfaction. Adolescent children demand more companionship and partnership in the face of difficulties. Decision/Commitment is linked to support, love and consideration for the spouse (Givertz, Segrin, & Woszidlo, 2016). This phase, called “companion love” by Hatfield, Pillemer, O’Brien, and Le (2008), involves the decision to keep the relationship above all problems (adolescence), based on the importance of the partner and the romantic relationship, and also for other reasons, such as child rearing, economic dependence, fear of loneliness or lack of strength to start over.

Finally, for men and women in the G4 group, the TLS-R Passion predictor variable contributed the most to explaining Marital Satisfaction. This factor of the Triangular Theory of Love relates to aspects of physical attraction and sexual contact, involving sexual desires, as well as behavioral, affective, and cognitive records of affection by the partner. According to Aboim (2009), contrary to marital dissolutions, the importance attached to both individual and marital well-being and the erotic aspect of Passion is identified as a benefit to marriage. However, Hatfield et al. (2008) reported that Passion, over the years, tends to diminish in the face of other feelings. Older adults reported lower levels of passion than younger adults did, however, this decrease is small (Ahmetoglu et al., 2010). Older women and men still reported moderate to high levels of passionate and erotic love (Grote & Frieze, 1998).

It was observed that the Passion of the Triangular Theory of Love could be divided into Erotic Passion and Romantic Passion (Gouveia, Carvalho, Santos, & Almeida, 2013). The Erotic Passion would be the experience of the desires and needs of a physiological nature. This component of love would reach its maximum level within half a year into the relationship. Then it could experience a dramatic decline and, after some stability, a slight decline for around four years. Concomitantly, the Romantic Passion, the idealization of the partner, the intrusive and continuous thoughts, the belief in the omnipotence of love, the desires and needs of love, would tend to grow gradually from the beginning of the relationship. The G4 male and female data from the current study could be reflecting this romantic part of Passion. The couple goes back to living alone when their children leave home to study, travel, live alone and raise their own families. There is a need to return to Passion to maintain a healthy relationship and marital satisfaction.

Jackson, Miller, Oka, and Henry (2014) conducted a meta-analysis of 226 independent samples which make up for over 100,000 participants to test the assumption that women experience less marital satisfaction than men and did not find significant differences between female and male individuals. In the present study, no differences were found between men and women (main effect) and the interaction between the family life cycle variables and sex (interactive effect) on marital satisfaction.

However, differences in levels of marital satisfaction were observed between the stages of the family cycle. The G1 group presented the highest RAS average levels followed by the G4, the G2 and, finally, the G3 group, with the lowest levels (Table 2). For Rios and Gomes (2009), most couples without children reported that this condition provides increased marital and personal satisfaction, given that the time that would be invested in children can be channeled to the dyad and to the individual, which would explain these higher RAS levels.

Lower levels of RAS scores in individuals with children could be explained by the interaction between marital and parental roles. Conjugality and parenting are systems, although distinct ones, which are interconnected. The more spouses feel part of a parental alliance, the greater the proximity and marital satisfaction (Augustin & Frizzo, 2015). Coparenting demands effective participation, responsibility, and interest in the creation of children. Undoubtedly, there are numerous challenges in the exercise of parenting that can lead to stress, conflict and new adaptations. In the phase of couples with adolescent children, there is a resurgence of difficulties in the exercise of the parental role. Adolescents often put themselves in the position of questioners of family functioning, since they have typical experiences of their age that can generate insecurity, confusion, anguish, ambivalent feelings, and misunderstanding. As a result, effective strategies for dealing with this critical phase are required from parents. According to Augustin and Frizzo (2015), the basic components of coparenting are coparental support versus depreciation, the individual agreement on parenting and the division of roles, and managing family interactions. The form of administration of these components will have repercussions on Marital Satisfaction. For Tissot and Falcke (2017), dyadic satisfaction presented the lowest level in individuals with young children and the highest score in individuals with adolescent children.

This study revealed that the TLS-R and the RBS love dimensions had a stronger impact on marital satisfaction in the early and late stages of the family life cycle (G1, G3, and G4). Aspects related to Social Skills had a predominant influence on Marital Satisfaction in the stage of couples with young children (G2), considered the most crucial one of the family life. These data can help couples anticipate crises by pursuing and participating in training programs.

It should be considered that the current results come from a convenience sample and therefore probably do not represent the Brazilian population. On the other hand, it is a sample composed entirely of married individuals, legally or not, and society in general. It is suggested that new investments may use dyadic data to explore possible factors associated with marital satisfaction through samples from different Brazilian regions. All of this, combined with longitudinal designs, would provide more consistent results in examining this variable throughout the family life cycle.