This is an open access article distributed under the terms of the Creative Commons
Attribution License (Introduction
Psychological stress has been conceptualized as an experience arising from continuous interactions and adjustments between a person and his environment (Lazarus & Folkman, 1984). It was also suggested that stress is not inherently maladaptive (Selye, 1973, 1974) and that, at moderate levels, it can foster positive changes in wellbeing, adaptability, and performance of individuals (Cotton & Hart, 2003; Hart & Cotton, 2002; McGowan, Gardner, & Fletcher, 2006; O’Sullivan, 2011; Quick, Nelson, & Quick, 1990; Tedeschi & Calhoun, 2004). In contrast, “when the demands placed on the body (in the larger sense that includes both the physiological and the psychological aspects) exceed its capacity to expend energy in maintaining homeostasis” (Le Fevre, Matheny, & Kolt, 2003, p. 729) a negative psychological response can occur.
In family studies, parenting stress has been conceptualized as a perceived discrepancy between situational demands and personal resources connected to parenthood (Abidin, 1995; Cooper, McLanahan, Meadows, & Brooks-Gunn, 2009; Deater-Deckard, 2004) and it is distinct from stress in other domains of life (e.g., work stress or marital relationship stress). Because high levels of stress that arises from the demands of parenthood has been shown to impact several areas of family life, including both the quality of parent-child relationship and child adjustment (Östberg & Hagekull, 2013), it is important to understand which factors may predict this parental condition. At this regard, the attention to the quality of marital functioning and coparenting relationship may be particularly valuable.
Marital Adjustment and the Quality of Parent-Child Relationship
Past literature has consistently shown a significant relation between marital functioning and the quality of parent-child relationships and parenting stress. A spillover hypothesis (Easterbrooks & Emde, 1988; Erel & Burman, 1995) has explained this association by suggesting that aspects - such as affect or behavior - of one relationship in a family can transfer to another, such as from the marital relationship to the parent-child relationship. This hypothesis assumes that parents experiencing deteriorated marital relationships may show more dysfunctional parenting due to a spillover of the couple’s overall distress (Emery, Hetherington, & DiLalla, 1984). A series of studies has empirically tested that marital satisfaction affects the quality of the caregiver-child relationship in terms of parental investment, responsivity, self-efficacy, and reduced hostility (Cox, Owen, Lewis, & Henderson, 1989; Floyd, Gilliom, & Costigan, 1998; Merrifield & Gamble, 2013; Sturge-Apple, Davies, & Cummings, 2006). Other studies have also demonstrated how a couple’s relationship characterized by warmth, complicity, emotional support, and greater conflict management skills is related to lower levels of parenting stress (Colpin, De Munter, Nys, & Vandemeulebroecke, 2000; Mulsow, Caldera, Pursley, Reifman, & Huston, 2002; Wieland & Baker, 2010) and greater caregiving skills (Lafontaine, Belanger, & Gagnon, 2009). Inversely, marital conflict was found to be associated with parents’ withdrawal from a parenting role and their engagement in dysfunctional interactions denoted by permissiveness, hostility, and tension (Almeida, Wethington, & Chandler, 1999; Fauchier & Margolin, 2004; Kerig, Cowan, & Cowan, 1993; Krishnakumar & Buehler, 2000; Margolin, Gordis, & Oliver, 2004).
Coparenting and the Quality of Parent-Child Relationship
Recently, to better understand the mechanisms through which dysfunctional relationships in the marital dyad affect parenting, scholars have identified coparenting as an important aspect of the interparental relationship that can explain the effects of marital functioning on parent-child relationships (Bonds & Gondoli, 2007; Floyd et al., 1998; Frosch, Mangelsdorf, & McHale, 2000; Ippolito Morrill, Hines, Mahmood, & Cordova, 2010; Margolin, Gordis, & John, 2001; McHale & Rasmussen, 1998). Coparenting is the component of marital relationships that pertains specifically to parenting together and it is through this relationship that parents negotiate their respective roles, responsibilities, and contributions to their children (Margolin et al., 2001; McHale, Kuersten-Hogan, Lauretti, & Rasmussen, 2000). This construct encompasses both the parents’ abilities to coordinate their activities and support one another as well as their not supportive responses or efforts to interfere with or undermine one another (Feinberg, Brown, & Kan, 2012; McHale, 1995). More specifically, the broad construct of coparenting includes supportive or hostile-competitive dimensions, as well as discrepancies in parental involvement (Belsky, Putnam, & Crnic, 1996; McHale, 1995, 1997), triangulation (Margolin et al., 2001; McHale, 2007), and parents’ perceptions of their parenting alliance (Abidin, 1992; Cohen & Weissman, 1984; Floyd et al., 1998). Parenting alliance, in particular, has been one of the most widely used operationalizations of coparenting and, according to Weissman and Cohen (1985), a sound coparenting alliance includes the following four characteristics: (a) both parents’ investment in the child; (b) valuing each other’s involvement with the child; (c) respect for each other’s judgment about child rearing; and (d) having a desire to communicate child-related information.
An important hypothesis presented in the literature considers the coparenting relationship more of a proximal determinant of the quality of parenting experiences than other features of marriage, outlining that the parent’s ability to coordinate their activities can mediate the effects of general marital quality on parenting experiences (Bonds & Gondoli, 2007; Floyd et al., 1998; Ippolito Morrill et al., 2010; Margolin et al., 2001). According to Grych and Fincham (1993, p. 61) “the widely used concept of domain specificity” provides one explanation for this greater proximity of the coparenting relationship to parenting: marital interactions concerning child rearing have a greater impact on parenting than other aspects of the marital relationship (Ippolito Morrill et al., 2010). In this way, solid parenting alliance represents a point of intersection between two family subsystems - precisely, the marital and parent-child relationships (Lindsey, Caldera, & Colwell, 2005). Several studies have directly supported this proposed hypothesis, suggesting that any direct effect between marital relations and parenting decreases or disappears after accounting for the mediating effect of parenting alliance (Bonds & Gondoli, 2007; Floyd et al., 1998; Ippolito Morrill et al., 2010; Margolin et al., 2001). The dimensions of parenting investigated in these studies are: parenting stress (Margolin et al., 2001); parenting practices (Ippolito Morrill et al., 2010; Margolin et al., 2001), perceived parenting competence (Floyd et al., 1998), and maternal warmth (Bonds & Gondoli, 2007). Specifically, in a cross-sectional study conducted on both parents of preschool children, Margolin, Gordis, and John (2001) examined the mediational effects of coparenting - represented by the dimensions of cooperation, triangulation, and conflict - on the relationships among marital conflict, parenting practices, and parenting stress. Results showed that coparenting functions as a link in the relationship between marital conflict and parenting both for mothers and fathers. The data were consistent with a mediational model, indicating that coparenting is an important mechanism by which marital relations affect parent-child relationships. Floyd, Gilliom, and Costigan (1998) conducted a longitudinal study on a sample of married and cohabiting parents of school-age children with mental retardation: the results showed the parenting alliance mediated the effects of marriage on parenting experiences in terms of perceived parenting competence and negative parent–child interactions. Yet, in a longitudinal study conducted on both parents of preadolescents, Ippolito Morrill, Hines, Mahmood, and Cordova (2010) have empirically tested two different models: the traditional indirect model (parenting alliance as a mediator of the relationship between marital health and parenting practices) and an alternative model (in which parenting alliance simultaneously predicts both marital relationships and parenting practices). Findings suggested that both the traditional indirect model and the alternative predictor model fit for both spouses. On the basis of these results, the parenting alliance could perform multiple and dynamic roles in the overall family system. Finally, in a sample of mothers of adolescents, Bonds and Gondoli (2007) showed that the parenting alliance - characterized by good teamwork, mutual support, and consistent perception of the child - mediates the relationship between marital adjustment and maternal warmth. The findings suggested that positive marital relations set the stage for a solid parenting alliance that then facilitates the positive emotional tone of parent-child interactions.
In the literature, an interesting alternative perspective argued that more positive qualities of one subsystem can buffer against negative qualities of the other dyadic subsystem (Erel & Burman, 1995). Therefore, it was also suggested that the quality of coparenting relationship could moderate the association between marital functioning and parenting experiences because interactive effects between these variables may occur as well (Feinberg, 2002; Merrifield & Gamble, 2013). Thinking about coparenting and parenting alliance in particular as moderator entails, on the one hand, to assume that there might be a good parenting alliance despite the marital distress (Cowan & McHale, 1996; Talbot & McHale, 2004) and, on the other hand, to believe that the strength of the relationship between marital adjustment and parenting experiences could be smaller for parents with high levels of a parenting alliance and greater for those with low levels. In fact, it may be that, at high levels of coparenting support and alliance, the parental ability to cooperate for the wellbeing of their children provides a protective basis for positive parent-child interactions, even in the presence of marital distress. To our knowledge, there is only one study by Merrifield and Gamble (2013), that directly assessed the interactional effects of marital quality and coparenting relationship on parenting experiences, measured in terms of self-efficacy. Authors explored the possibility that supportive coparenting may compensate for the potentially damaging effects of less satisfaction in a marriage, or more spousal conflict on parenting self-efficacy. Alternatively, positive marital relations may compensate for the potentially negative effects of undermining coparenting on parenting cognitions. Findings show mixed evidence for a stress buffering association. Authors showed that mothers and fathers’ reports of supportive coparenting were associated with increases in parenting self-efficacy in the context of more positive marital relationships. Additionally, fathers’ reports of higher marital satisfaction were associated with a greater decrease in parenting self-efficacy, in the context of reports of greater undermining coparenting.
In literature, further works that evaluated the effects of interaction between marital conflict and coparenting quality on child adjustment suggested that a strong parenting alliance may diminish the effects of couple conflict on children (Abidin & Brunner, 1995; McHale, 1995). Specifically, high levels of couple hostility may be detrimental only when coparenting quality is low, not when it is high.
The Present Study
This review of the literature suggests that marriage and parenting subsystems in families could be connected through mediational processes. In this study, these processes will be explored in an attempt to understand if parenting alliance could be considered the mechanism through which marital adjustment influences the parenting stress.
Moreover, this study is aimed to extend this line of exploration by focusing on both mothers and fathers, and by exploring the effects of the specific dimensions of marital adjustment on both maternal and paternal stress. The choice to focus on both mothers’ and fathers’ reports regarding marriage, parenting alliance and parenting stress, starts form the suggestions of the literature that outlined how this research topic could gain in importance as researchers increasingly focus on both mothers’ and fathers’ roles in the family system (Bronte-Tinkew, Horowitz, & Carrano, 2010; Feinberg, 2002). At this regard, we can say that the majority of the studies on parenting stress have investigated maternal stress and, when research has focused on fathers, it has generally relied on information reported by mothers about fathers (Bronte-Tinkew et al., 2010).
Unlike most studies on parenting stress, those on coparenting, has considered both mothers’ and fathers’ roles, without detecting significant differences because of parent gender (Floyd et al., 1998; Ippolito Morrill et al., 2010; Margolin et al., 2001; Merrifield & Gamble, 2013). These data, although consistent with recent research that suggests the influence of gender on parenting roles is decreasing in importance (Coltrane & Adams, 2008), deviate from those of research studies that found that fathers display more marriage-related disruptions in parenting than do mothers, suggesting that fathers’ parenting may be more vulnerable (the father vulnerability hypothesis) to negative marital relations (Belsky, Gilstrap, & Rovine, 1984; Cummings, Goeke-Morey, & Raymond, 2004; Cummings & O’Reilly, 1997; Parke, 2002). We suppose that it’s therefore interesting to explore both mothers and fathers reports regarding the associations among marital adjustment, parenting alliance, and parenting stress. In exploring these associations, we also interested to consider the distinct dimensions of marital adjustment (dyadic consensus, dyadic satisfaction, dyadic cohesion, and affectional expression), measured through the Dyadic Adjustment Scale (DAS; Spanier, 1976) that is the most widely utilized self-report measure of relationship adjustment in the social and behavioral sciences (South, Krueger, & Iacono, 2009).
Moreover, we are interested in exploring the associations among marital adjustment, parenting alliance, and parenting stress in families with school-aged children. We chose to study this specific age group because previous studies have investigated this topic in families with preschool children and adolescents, to our knowledge, only Floyd et al. (1998) have examined school-aged children with developmental disabilities. In addition, the focus on this age group stems from the consideration that the elementary school children require specific skills and tasks by the parents. From a family system perspective, Roberts (1990) stated that the main parenting tasks are establishing rules which govern interactions, maintaining appropriate hierarchy, and allowing the child to establish relationships outside the home. These parenting tasks depends on the closeness/distance among parents (Broderick & Smith, 1979) and are supported by their ability to maintain good coparenting competences.
Objectives and Hypothesis
The primary goal of the present study was to examine the role of parenting alliance as a mediator of the relation between marital adjustment and maternal and paternal stress. More precisely, we were interested in exploring which specific dimensions of marital quality - in terms of dyadic consensus, dyadic satisfaction, dyadic cohesion, and affectional expression - are mediated by parenting alliance. In fact, the effects of these specific dimensions were not explored in previous literature on parenting stress. Further, to explore differences based on parents’ gender, the present study examined parenting alliance as a mediator - separately for mothers and fathers - between marital adjustment and maternal and paternal stress in a sample of Italian families with school-aged children (6-11 years).
We examined the hypothesis stating that marital adjustment (in terms of dyadic consensus, dyadic satisfaction, dyadic cohesion, and affectional expression) leads to increased perceived parenting alliance of both mothers and fathers, which, in turn, leads to reduced maternal and parental stress. On the basis of fathering-vulnerability hypothesis, we expect the parenting stress of fathers could be more susceptible to deterioration in the face of marital disadjustment.
Our secondary goal was to explore if mothers and fathers’ parenting alliance could moderate the relation between marital adjustment and maternal and paternal stress. The consideration of parenting alliance as a potential protective factor in the relation between marital relationships and parenting stress is an attempt to move beyond the prevalent mediational hypothesis in order to investigate if parenting alliance could play other important possible roles within the family system (Feinberg, 2002; Merrifield & Gamble, 2013).
Methods
Participants
Participants were 236 mothers and 236 fathers of Italian pupils (49.2% boys, 50.8% girls) ages 6-11 years (M = 8.6; SD = 1.7) recruited by four primary public schools located in Milan and in the Province of Milan. The children were noted as being the only child (36.8%), the first-born (32.9%), the second-born (23.7%), and the third-born (6.6%). The couples had been married 12 years on average (SD = 4.5). The mothers averaged 40.9 years of age (SD = 4.4), and the fathers averaged 42.9 years of age (SD = 4.8). Socioeconomic status (SES) of participants’ families was assessed by asking for parents’ qualifications and jobs: 32% of participants were from middle-low class, 47.8% from middle class, and 20.2% from middle-upper class.
Procedure
Participants were recruited from four primary public schools located in Milan or the Province of Milan. They are part of a broader sample recruited in a previous research aimed to validate the Parenting Alliance Measure with the Italian population (Camisasca, Miragoli, Caravita, & Di Blasio, in press). The participating schools were recruited by a standard procedure that included introductory meetings with school principals and letters to the parents describing the goals and procedures of the study. Two hundred fifty-five dyads married or cohabitant parents indicated their willingness to participate in this study. These parents signed consent forms that described the project and its goals, the voluntary nature of participation, and the confidentiality of the data collected. Packets consisting of self-report measures were delivered to parents, asking them to complete the forms independently and not to share their answers. Mothers and fathers were asked to complete three questionnaires: the Dyadic Adjustment Scale (DAS; Spanier, 1976; Italian validation by Gentili, Contreras, Cassaniti, & D’Arista, 2002), the Parenting Alliance Measure (PAM; Abidin & Konold, 1999), and the Parenting Stress Index Short Form (PSI-SF; Abidin, 1995; Italian validation by Guarino, Di Blasio, D’Alessio, Camisasca, & Serantoni, 2008) to be returned within two months. Of the 255 dyads of parents, 19 (7.5%) were excluded because their scores were below the 10th percentile of the defensiveness scale of PSI-SF, which taps the extent to which parents responded with a set that attempts to present themselves and their relationship with their child in an overly positive manner. The present analyses were based on 236 mothers and 236 fathers who completed all instruments.
Measures
Dyadic Adjustment Scale (DAS; Spanier, 1976; Italian validation by Gentili et al., 2002) is a widely used 32-item self-report measure of the quality of the marital relationship consisting of four subscales: (1) dyadic consensus (13 items), the degree to which the couple agrees on matters of importance to the relationship (e.g., religion, recreation, friends, household tasks, and time spent together); (2) dyadic satisfaction (10 items), which evaluates the degree to which the couple is satisfied with their relationship, expressed in the frequency of quarrels, discussions of separation and positive interactions; (3) affectional expression (4 items), which evaluates the degree of demonstrations of affection and the frequency of disagreements regarding expressed levels of affection and sex; and (4) dyadic cohesion (5 items), which evaluates the degree of closeness and the frequency of shared activities experienced by the couple. Items’ ratings vary with rating of agreements ranging from 0 (Always Disagree) to 5 (Always Agree), rating of frequency from 0 (All the Time) to 5 (Never) or from 0 (None) to 4 (All), dichotomous ratings ranging from 0 (Yes) to 1 (No), and qualitative ratings ranging from 0 (Extremely Unhappy) to 6 (Perfect). The 32 items are summed to create a total score ranging from 0 to 151, with higher scores indicating better marital adjustment.
The values of internal consistency of the Italian validation of the DAS (Gentili et al., 2002) correspond to: α = .93 for the total adjustment; α = .89 for the dyadic consensus; α = .87 for the dyadic satisfaction; α =.63 for the affective expression; and α = .78 for the dyadic cohesion. In our sample: α = .90 for the total adjustment scale (α = .90 for mothers and fathers); dyadic consensus: α = .86 (mothers) and α = .87 (fathers); dyadic satisfaction: α = .73 (mothers) and α = .71(fathers); affective expression: α = .69 (mothers) and α = .72 (fathers); and dyadic cohesion: α = .80 (mothers) and α = .76 (fathers).
In some studies (Sabourin, Valois, & Lussier, 2005; Sharpley & Cross, 1982) this four-factor structure of the DAS (Spanier, 1976) was not replicated as a one underlying "adjustment" dimension emerged. However, other researchers have confirmed the multi-dimensional nature of the construct which the four-factors DAS was reported to measure adequately (South et al., 2009). In order to examine the dimensionality of DAS, in our sample, a confirmatory factor analysis was carried out by EQS (Bentler, 2001). We compared the fit indexes of two CFA models: a first model in which the subscale scores have been specified as loading a unique factor, according to Sharpley and Cross (1982) and Sabourin et al. (2005), and a second CFA model in which the subscale scores were loading the four-factor model introduced by Spanier (1976) and replicated by South, Krueger, and Iacono (2009). For the one-factor model, goodness-of-fit indexes were not satisfactory (χ2(464) = 1483.16, p < .001; CFI =. 72; RMSEA = .08; SRMR = .070; 90% CI [.080, .089]; χ2/df = 3.19). On the contrary, the confirmatory factor analysis applied on the four factors version showed fit indexes (χ2(455) = 1151.02, p < .001; CFI = .82; RMSEA = .07; SRMR = .070; 90% CI [.065, .075]; χ2/df = 2.52) better than those of the one-factor model.
Parenting Alliance Measure (PAM; Abidin & Konold, 1999; Italian validation Camisasca et al., in press) is a self-report measure assessing the degree to which parents believe they have a sound parenting relationship with their child’s other parent. PAM consists of 20 items that can be responded independently by either mothers or fathers of children from 1 to 19 years old. The item response scale is a 5-point Likert scale, ranging from 1 (Strongly Disagree) to 5 (Strongly Agree), with higher scores reflecting stronger co-parental alliance.
PAM was developed by Abidin and Konold (1999) on the basis of Weissman and Cohen’s (1985) four characteristics of a sound parenting alliance (see the section Introduction). PAM has been normed over large samples in the North-American population, and it has been found to measure the same factors for fathers and mothers (Konold & Abidin, 2001). In the original validations sample the internal consistency of PAM was high: α = .97 for mothers, α = .96 for fathers, and α = .97 for a combined sample (Abidin & Konold, 1999). The Italian version of the PAM was translated from the Parenting Alliance Measure by Abidin and Konold (1999) with the special permission of the Publisher, Psychological Assessment Resources, PAR, Inc. In our contribute to validation (Camisasca et al., in press) total score alphas were high for both mothers (.90) and fathers (.92).
Parenting Stress Index Short Form (PSI-SF; Abidin, 1995; Italian validation by Guarino et al., 2008) is a 36-item questionnaire that assesses parenting stress. The items are rated on a 5-point scale, ranging from “Strongly Agree” to “Strongly Disagree.” It consists of four subscales: parental distress (PD), 12 items; difficult child (DC), 12 items; and parent-child dysfunctional interaction (P-CDI), 12 items, as well as a defensive responding subscale that consists of 7 items drawn from the parental distress subscale. The defensive responding subscale assesses parental bias in reporting by quantifying the desire of parents to present a favorable impression of themselves and minimize problems in the parent–child relationship. The PD subscale focuses on the sense of competence/incompetence in rearing the child, conflict with the partner, lack of social support, and stress associated with the restrictions deriving from the role of parent. The DC subscale focuses the parent’s perception of the child in terms of temperament, requesting and provoking behaviors, and non-collaborative and demanding behaviors. Finally, the P-CDI subscale measures parents’ perceptions of the emotional quality of their relationship with their children. The sum of the scores of the three subscales (PD + P-CDI + DC) enables us to obtain the value of total stress, which gives an indication of the overall level of the specific parental stress, not deriving from other roles or other events. The values of internal consistency of the Italian validation of the PSI-SF (Guarino et al., 2008) correspond to α = .91 for the total stress scale; α = .91 for the PD subscale; α = .95 for the P-CDI subscale, and α = .90 for the DC subscale. In this study, regarding to the mothers, the values of internal consistency of the PSI-SF correspond to α = .90 for the total stress scale; α = .81 for the PD subscale; α = .83 for the P-CDI subscale, and α = .81 for the DC subscale. Regarding to the fathers, the values of internal consistency of the PSI-SF correspond to α = .91 for the total stress scale; α = .85 for the PD subscale; α = .83 for the P-CDI subscale, and α = .82 for the DC subscale.
Strategy of Analysis
Data analyses proceeded in several steps. First, correlations were calculated among all variables investigated (four dimensions of marital adjustment, parenting alliance, and parenting stress) to examine initial bivariate associations and identify possible covariates.
Second, a set of regression analyses were conducted to examine whether parenting alliance mediates the relationship between marital adjustment and parenting stress. Four regression parameters were estimated separately for mothers and fathers, based on procedures recommended by Baron and Kenny (1986): (a) the predictor (marital adjustment) must be significantly associated with the hypothesized mediator (parenting alliance); (b) the predictor must be significantly associated with the dependent measure (parenting stress); (c) the mediator must be significantly associated with the dependent variable, and (d) the impact of the predictor on the dependent measure is less after controlling for the mediator. The procedures outlined by Baron and Kenny (1986) were deemed appropriate, as mediation analysis in psychological research is most often guided by their criteria. Finally, a set of regression analyses were conducted to examine the possibility of moderating effects of marital adjustment and parenting alliance on parenting stress. Separately for mothers and fathers, we computed two-step hierarchical multiple regression analyses (Baron & Kenny, 1986; Holmbeck, 1997), using the centered four dimensions of marital adjustment and parenting alliance in Step 1 and the same centered variables and the interaction products among these variables in Step 2.
Results
Associations Among Marital Adjustment, Parenting Alliance and Parenting Stress
Correlations, means, and standard deviations of all variables used in the present study are presented in Table 1 for mothers and in Table 2 for fathers.
Regarding the variables investigated, the means scores are similar to those obtained in the Italian validation studies (DAS: Gentili et al., 2002; PAM: Camisasca et al., in press; PSI: Guarino et al., 2008) both for mothers and fathers, and are placed within normal limits. More precisely, by considering the DAS Total score, our data show that 49.6% of mother and 47% of fathers have a score lower than the median, that is 117 for mothers and 119 for fathers. Regarding parenting alliance, our data show that 49.2% of mother and 47.9% of fathers have a score lower than the median, that is 85 for mothers and 84 for fathers. Finally, regarding parenting stress, our data show that 52.1% of mother and 47.5% of fathers have a score lower than the median, that is 70 for mothers and 67 for fathers.
Table 1
Descriptive Statistics and Correlations for Marital Adjustment (DAS), Parenting Alliance (PAM), and Parenting Stress (PSI-SF) in Mothers (N = 236)
| 1 | 2 | 3 | 4 | 5 | 6 | M (SD) | |
|---|---|---|---|---|---|---|---|
| 1. Dyadic consensus (DAS) | - | 53.7 (6.0) | |||||
| 2. Dyadic satisfacton (DAS) | .353** | - | 35.9 (6.7) | ||||
| 3. Affectional expression (DAS) | .529** | .334** | - | 9.3 (2.0) | |||
| 4. Dyadic cohesion (DAS) | .498** | .276** | .396** | - | 16.5 (4.9) | ||
| 5. Parenting alliance (PAM) | .590** | .266** | .431** | .421** | - | 84.1 (9.3) | |
| 6. Parenting stress (PSI-SF) | -.320** | -.186** | -.281** | -.244** | -.384** | - | 70.3 (15.9) |
*p < .05. **p < .01.
Table 2
Descriptive Statistics and Correlations for Marital Adjustment (DAS), Parenting Alliance (PAM), and Parenting Stress (PSI-SF) in Fathers (N = 236)
| 1 | 2 | 3 | 4 | 5 | 6 | M (SD) | |
|---|---|---|---|---|---|---|---|
| 1. Dyadic consensus (DAS) | - | 53.8 (6.1) | |||||
| 2. Dyadic satisfacton (DAS) | .379** | - | 37.4 (6.1) | ||||
| 3. Affectional expression (DAS) | .539** | .339** | - | 9.4 (2.2) | |||
| 4. Dyadic cohesion (DAS) | .484** | .358** | .390** | - | 16.3 (4.7) | ||
| 5. Parenting alliance (PAM) | .411** | .431** | .320** | .291** | - | 84.7 (9.2) | |
| 6. Parenting stress (PSI-SF) | -.265** | -.290** | -.211** | -.239** | -.360** | - | 66.7 (15.9) |
*p < .05. **p < .01.
For mothers and fathers, the correlation analysis showed that the variables were intercorrelated. In particular, maternal perception of parenting alliance was correlated positively with all dimensions of marital adjustment (dyadic consensus: r = .59, p < .001; dyadic satisfaction: r = .27, p < .001; affectional expression: r = .43, p < .001; dyadic cohesion: r = .42, p < .001) and negatively with parenting stress (r = -.38, p < .001). Similarly, paternal perception of parenting alliance was correlated positively with all dimensions of marital adjustment (dyadic consensus: r = .41, p < .001; dyadic satisfaction: r = .43, p < .001; affectional expression: r = .32, p < .001; dyadic cohesion: r = .29, p < .001) and negatively with parenting stress (r = -.36, p < .001).
Mediating Effects of Parenting Alliance in Mothers and Fathers
To test the effects of the four dimensions of marital adjustment separately for mothers and fathers (predictors: dyadic consensus, dyadic satisfaction, affectional expression, and dyadic cohesion) on parenting alliance (mediator), multiple linear regressions were performed (Condition 2 of the mediational model). To examine the relative contributions of marital quality (predictor) and parenting alliance (mediator) to parenting stress (outcome), the hierarchical multiple regression analyses were conducted separately for mothers and fathers. Marital adjustment’s dimensions - dyadic consensus, dyadic satisfaction, affectional expression, and dyadic cohesion - were entered in Step 1 (Condition 1 of the mediation model). The parenting alliance was added in Step 2 (Condition 3) to determine what degree it mediates the relationship between marital adjustment and parenting stress (Condition 4).
Table 3 presents the results of regression analysis (unstandardized and standardized coefficients) predicting association between marital adjustment (dyadic consensus, dyadic satisfaction, affectional expression, and dyadic cohesion) and parenting alliance, and Table 4 the results of the hierarchical multiple regression analysis predicting parenting stress.
Table 3
Predictors of Parenting Alliance in Mothers (N = 236)
| Predictor variables | B | SE B | β |
|---|---|---|---|
| Constant | 35.85 | 4.53 | |
| Dyadic consensus | .68 | .10 | .44** |
| Dyadic satisfaction | .04 | .08 | .03 |
| Affectional expression | .62 | .29 | .13* |
| Dyadic cohesion | .27 | .12 | .14* |
Note. R2 = .38; F = 35.99**.
*p < .05. **p < .01.
Table 4
Hierarchical Regression Analysis Predicting Parenting Stress in Mothers (N = 236) and Fathers (N = 236)
| Predictor | Mother
|
Father
|
||
|---|---|---|---|---|
| ∆R2 | β | ∆R2 | β | |
| Step 1 | .15 | .12 | ||
| Dyadic consensus | -.15* | -.12 | ||
| Dyadic satisfaction | -.04 | -.20** | ||
| Affectional expression | -.11 | -.40 | ||
| Dyadic cohesion | -.20** | -.09 | ||
| Step 2 | .02 | .05 | ||
| Dyadic consensus | -.07 | -.06 | ||
| Dyadic satisfaction | -.04 | -.12 | ||
| Affectional expression | -.09 | -.02 | ||
| Dyadic cohesion | -.17* | -.08 | ||
| Parenting alliance | -.19* | -.25** | ||
| Total R2 | .17 | .17 | ||
*p < .05. **p < .01.
Regarding the mediating role of maternal perception of parenting alliance, results revealed that the parenting alliance (β = -.19, t = -2.52, p < .05) totally mediates the relationship between the dyadic consensus (β = -.07, t = -.77, p = n.s.) and parenting stress and partially mediates the relation between dyadic cohesion (β = -.17, t = -2.34, p < .05) and parenting stress. The value for the Sobel test (Sobel, 1982, 1986) was indeed significant (t = -.163, p = .05). Figure 1 shows a schematic of the mediation model under investigation.
Figure 1
Parenting alliance as mediator of the relationship between marital adjustment and parenting stress in Mothers (N = 236).
Table 5 presents the results of regression analysis (unstandardized and standardized coefficients) predicting interaction between marital adjustment (dyadic consensus, dyadic satisfaction, affectional expression, and dyadic cohesion) and parenting alliance, and Table 4 gives the results of the hierarchical multiple regression analysis predicting parenting stress.
Table 5
Predictors of Parenting Alliance in Fathers (N = 236)
| Predictor variables | B | SE | β |
|---|---|---|---|
| Constant | 44.02 | 4.98 | |
| Dyadic consensus | .36 | .11 | .24** |
| Dyadic satisfaction | .46 | .10 | .30** |
| Affectional expression | .31 | .29 | .07 |
| Dyadic cohesion | .08 | .13 | .04 |
Note. R2 = .26; F = 20.57**
*p < .05. **p < .01.
Regarding the mediating role of parenting alliance in fathers (see Table 4), results revealed that parenting alliance (β = -.25, t = -3.58, p < .001), totally mediates the relationship between dyadic satisfaction (β = -.12, t = -1.70, p = n.s.) and parenting stress. Figure 2 shows a schematic of the mediation model under investigation.
Figure 2
Parenting alliance as mediator of the relationship between marital adjustment and parenting stress in fathers (N = 236).
Moderating Effects of Marital Adjustment and Parenting Alliance in Mothers and Fathers
To explore the moderating effects of marital adjustment and parenting alliance on parenting stress separately for mothers and fathers, we computed a two-step hierarchical regression analyses, according to Baron and Kenny’s (1986) recommendations for testing moderator effects. The analyses assessed the unique effects of marital adjustment (in term of dyadic consensus, dyadic satisfaction, affectional expression, and dyadic cohesion) and parenting alliance, and the association of the interaction of the four dimensions of marital adjustment and parenting alliance with parenting stress. Two-way interactions between each measure of marital adjustment and parenting alliance assessed whether the parenting alliance moderated the relationship of marital adjustment with parenting stress. Evidence of moderation would be apparent if there was a statistically significant change in the variance accounted for parenting stress with the introduction of the interaction terms (Cohen & Cohen, 1983).
The main effects of the four dimensions of marital adjustment and parenting alliance were entered at Step 1 to assess their unique association with the outcome measures. These main effects and the two-way interaction between each measure of marital adjustment and parenting alliance were entered at Step 2. Interaction terms were composed of the product of the variables (dyadic consensus, dyadic satisfaction, affectional expression, and dyadic cohesion X parenting alliance) that were centered to avoid problems of multicollinearity.
Table 6 reports standardized coefficients for the hierarchical regression analyses predicting maternal and paternal parenting stress. Regarding the mothers, results showed that at the Step 1, the main effects of centered parenting alliance (β = -.19, t = -2.52, p < .05) and centered dyadic cohesion (β = -.17, t = -2.34, p < .05) were significant, but at the Step 2, we did not detect a statistically significant change in the variance explained by the model (∆R2 = .02, ∆F = .32). Therefore, for the mothers, no moderating effects of marital adjustment and parenting alliance on parenting stress were sustained.
Table 6
Hierarchical Regression Analysis Predicting Parenting Stress in Mothers (N = 236) and Fathers (N = 236)
| Predictor | Mother
|
Father
|
||
|---|---|---|---|---|
| ∆R2 | β | ∆R2 | β | |
| Step 1 | .17 | .17 | ||
| Dyadic consensus | -.07 | -.06 | ||
| Dyadic satisfaction | -.04 | -.12 | ||
| Affectional expression | -.09 | -.02 | ||
| Dyadic cohesion | -.17* | -.08 | ||
| Parenting alliance | -.19* | -.25** | ||
| Step 2 | .02 | .02 | ||
| Dyadic consensus | -.06 | -.01 | ||
| Dyadic satisfaction | -.05 | -.13 | ||
| Affectional expression | -.09 | -.05 | ||
| Dyadic cohesion | -.17* | -.09 | ||
| Parenting alliance | -.18* | -.22** | ||
| Dyadic consensus × Parenting alliance | -.07 | -.17 | ||
| Dyadic satisfaction × Parenting alliance | .17* | .07 | ||
| Affectional expression × Parenting alliance | -.10 | -.06 | ||
| Dyadic cohesion × Parenting alliance | .02 | .01 | ||
| Total R2 | .19 | .19 | ||
*p < .05. **p < .01.
Likewise, regarding fathers results showed that at the Step 1, the main effect of centered parenting alliance (β = -.25, t = -3.58, p < .001) was significant and, at the Step 2, we did not detect a statistically significant change in the variance explained by the model (∆R2 = .01, ∆F = .52), showing that parenting alliance did not moderate any dimensions of marital quality on parenting stress.
Discussion
In this study, we explored the role of parenting alliance as a mediator or a moderator of the relation between marital adjustment and parenting stress. Our results support the hypothesis that parenting alliance could function as a link in the relationship between marital adjustment and parenting stress for both mothers and fathers but did not sustain the moderating role of parenting alliance for both parents. Specifically, our data are consistent with a mediational model, indicating that parenting alliance could be an important mechanism by which marital adjustment affect parent-child relationships.
In other words, marital adjustment could foster a solid parenting alliance that is characterized by good teamwork, mutual support, and consistent perceptions of the child, which then could facilitate the positive parent-child interactions and reduce levels of parenting stress. In this way, we can assume that a solid parenting alliance could provide the parents with emotional and psychological resources required to cope with parenting demands without stress. Our findings are consistent with results of both longitudinal and cross-sectional studies (Bonds & Gondoli, 2007; Floyd et al., 1998; Ippolito Morrill et al., 2010; Margolin et al., 2001) that any direct effect between marriages and parenting decreases or disappears after accounting for the mediating effect of coparenting.
Differently from more recent research (Ippolito Morrill et al., 2010) that did not detect significant differences in parenting alliance’s mediational effects because of parent gender, and differently from the fathering-vulnerable hypothesis, our findings show that mothers and fathers are sensitive to different dimensions of marital quality. More precisely, our data suggested that perceived mothers and fathers’ parenting alliance may function as a link in the relationship among different dimensions of marital adjustment and parenting stress. In particular, in the case of mothers, the perceived parenting alliance mediates the relationship among the dimensions dyadic consensus and dyadic cohesion, and the parenting stress. On the other hand, in case of fathers, results pointed out that parenting alliance mediates the relationship between dyadic satisfaction and parenting stress. In other words, mothers’ perceptions about their agreement with the partner on matters of importance to the relationship and mothers’ perceptions about closeness and the frequency of shared activities experienced by the couple are the most relevant dimensions of marital adjustment that affect parenting alliance, which then leads to a reduced maternal stress. On the contrary, for fathers, dyadic satisfaction - expressed in the frequency of positive interactions, few quarrels, or discussions of separation - is the most relevant dimension of marital adjustment that affects the parenting alliance, which then leads to a reduced paternal stress.
These results lead us to believe that, for mothers, aspects of the couple relationship (consensus and cohesion) that have the greatest influence on the parent-child relationship are closely associated with the possibility of being able to cooperate with partners, sharing both the value-related aspects of the philosophy of family life and the practical management of everyday life. Fathers seem more sensitive to the level of satisfaction derived from the emotional relationship with their partner and less receptive to the issues of cooperation about the practical management and the value-related aspects of family life. We could tentatively explain these differences assuming that mothers have generally more family responsibilities and household tasks as parents, independently of the functioning of their marital relationship. As a consequence, mothers could be more sensitive than fathers to the aspects of couple relationship that includes the possibility of sharing and feeling supported by their partners in family values, aims, and tasks. On the contrary, fathers are generally less involved in the management of everyday life and, thus, seem to be more sensitive to the quality of emotional relationship with the partner than to other aspects of marital adjustment.
The idea that mothers could be more receptive than fathers to the issues concerning the practical management of home environment and that fathers could be more sensitive than mothers to the level of marital satisfaction has also been suggested in one study by Nelson, O’Brien, Blankson, Calkins, and Keane (2009). This study was aimed to explore relations between four sources of family stress (marital dissatisfaction, home chaos, parental depressive symptoms, and job role dissatisfaction) and parenting experiences of both mothers and fathers. Results have confirmed that in case of fathers marital dissatisfaction was predictive of their parenting, while in case of mothers, the perception of the home environment as disorganized (home chaos) predicted a greater negativity and reduced patience, during interactions with their children.
Since this is the first study that evaluates the mediating role of parenting alliance on the associations among the specific dimensions of marital adjustment on parenting stress, further studies are needed to confirm these differences in mothers and fathers.
Finally, results do not support the idea that perceived mothers’ and fathers’ parenting alliance could moderate the relation between marital adjustment and parenting stress. More specifically, results showed that parenting alliance does not affect the strength of the relation among the dimensions of marital relations and parenting stress. In this sense, the quality of parenting alliance is not a protective factor able to mitigate the effects of marital disadjustment on parenting stress. In conclusion, although cross-sectional, our data support the idea that the parenting alliance could be an important mechanism by which marital relations affect parent- child relations and not a variable that affects the direction of this relationship.
Limitations of the Study and Future Directions
This study includes some methodological shortcomings. The nature of our cross-sectional data cannot explicitly identify directional effects or causal links, while longitudinal data could provide stronger evidence of directionality or causality. As noted by Baron and Kenny (1986), the causal sequence generally described in a mediational model cannot be proved on the basis of cross-sectional data. Without measuring marital adjustment, parenting alliance, and parenting stress across time, this study cannot draw conclusions about directional effects. However our results are consistent with the literature (Bonds & Gondoli, 2007; Floyd et al., 1998; Ippolito Morrill et al., 2010; Margolin et al., 2001) that found that marital relations affect parenting alliance, which then facilitates positive parent-child relations.
Another limitation of the study is the use of only self-report data. Future research should use a multi-method approach, including observational methods, self-reports, and interviews in which parents can describe their marital and coparenting relationships and parenting stress. In addition, our sample was composed of Italian parents who were predominantly well-educated and middle-class; replications of our findings with a more heterogeneous sample would foster generalization of findings to a broader population.
Finally, a further and important limitation is the method of analysis of our dyadic data, by considering them as independent and bidirectional, and by ignoring the mutual interdependence of scores of the two partners. A good corrective measure of this limitation could be the use of Actor-Partner Interdependence Model (APIM) which integrates a conceptual view of interdependence in two person relationships with the appropriate statistical techniques for measuring and testing it (Cook & Kenny, 2005). APIM uses the parent dyad as the unit of analysis (Kashy & Kenny, 2000; Kenny, 1996; Kenny & Cook, 1999) and implies that the two members of the dyad influence each other in the form of partner effects, creating interdependence between members (Ledermann & Macho, 2009).
In future research we could use API mediation model (APIMeM) in order to test our mediational hypothesis and to assess the effects of couple interdependence on the variables investigated (marital adjustment, parenting alliance and parenting stress).
Despite certain limitations, the findings of the present study draw attention to the construct of parenting alliance and suggest that the parenting alliance could be one of the explicative mechanisms of the connection between marital and parent-child relations for both mothers and fathers. Our results also show that parenting alliance mediates, in case of mothers and fathers, the effects of different dimensions of marital quality on parenting stress.
Although our results did not detect any moderation effects of parenting alliance, we suggest that future research continue to explore the potential buffering role of parenting alliance. For example, it could be useful to investigate if parenting alliance moderates the association among the four dimensions of marital adjustment (dyadic consensus, dyadic satisfaction, affectional expression, and dyadic cohesion) and the specific dimensions of parenting stress to better understand if and how at high/low parenting alliance conditions, marital disadjustment differently affects maternal and paternal stress in terms of parental distress, parent-child dysfunctional interactions, and perceptions of the child as difficult and problematic. A more detailed understanding of the mechanism that explains associations among different family subsystems - couples, coparenting and parent-child relationships -could lead both to the implementation of preventive programs and the improvement of more effective clinical interventions aimed to help distressed couples with children.
Clinical Implications
Our results also raise important practical implications for clinical interventions for couples with children. According to literature (Bonds & Gondoli, 2007; Feinberg, 2002; Floyd et al., 1998), we can assert that, on the one hand, our findings suggest that marital therapy that helps spouses function better as a couple may be an important factor in improving parent–child relationships. On the other hand, since the parenting alliance mediates the association between marital adjustment and parenting stress for both mothers and fathers, this specific family subsystem might be targeted by intervention programs aimed to interrupt the link between marital disadjustment and dysfunctional parenting. Further, because some couples who experience difficulty as coparents may not label themselves as maritally distressed and could be more resistant to marital interventions, the opportunity to receive coparenting intervention may be a more acceptable offer than marital therapy, per se. In strategic and practical terms, it may be easier to gain parents’ involvement about the treatment goals if the intervention is focused upon parenting alliance. Such an intervention, aimed at fostering the wellbeing of the child, could be more accepted than a therapy focused on improving the couple relationship. Furthermore, the enhancement of a positive parenting alliance may be an incentive for repairing deteriorated marital relationships and preventing a more serious escalation in marital conflict. Therefore, clinical interventions focused on parenting alliance may be a good preventing strategy to mitigate spillover of distress from the marital to parent–child relationships.