Although the rates of teenage pregnancy in the United States have continuously declined in the past 20 years, they remain higher compared to rates in other industrialized nations, and continue to have considerable social and economic implications for adolescent parents and their children, as well as society at large (Hamilton et al. 2015). Birth rates tend to be greater among ethnic minority couples compared to non-Hispanic White adolescents. For example, in 2013, childbearing rates for both Hispanic and African American women between the ages of 15 and 20 years were about twice as high as the rate of their White counterparts (Hamilton et al. 2015). Notably, African American and Hispanic adolescents accounted for 57% of teen births that year. The difficulties associated with young parenthood, such as financial strain and relationship instability, may be magnified for these adolescent ethnic minority parents due to increased risk for poverty, lower rates of educational attainment, and a lack of psychological resources (Beers and Hollo 2009; Reidenbach and Weller 2010). Individually and in combination with one another, these factors may contribute to higher levels of parenting stress and increased conflict between young, minority mothers and fathers.

The important task of navigating the coparenting relationship, in which parents coordinate with each other to raise their child, may be difficult for ethnic minority mothers and fathers who are new to parenthood. Conceptually, the coparenting relationship and parents’ relationship with one another are distinct in that the former is motivated by concern for the child, whereas the latter is motivated by concern for the self and/or partner (Margolin et al. 2001; McHale 2007). Research suggests that the relationship between the couple (regardless of romantic or marital status) and the coparenting relationship are two independent but related couple/family processes (McHale and Lindahl 2011; Van Egeren 2003). That is, among married couples, the coparenting relationship is thought to emerge from the marital relationship that existed before childbirth; at the same time, this relationship possesses unique traits and functions associated with the couple’s ability to rear their children together.

Theoretical considerations and empirical studies in recent decades have demonstrated the multidimensionality of coparenting (Feinberg 2003; Margolin et al. 2001). Teubert and Pinquart (2010) argue that cooperation, childrearing agreement, conflict, and triangulation comprise significant aspects of this construct. In line with these theoretical dimensions, coparenting has been defined as how parents communicate, manage conflict, make decisions, and collaborate with each other about raising their child (McHale and Lindahl 2011). Traditionally, the quality of coparenting has been conceptualized in terms of the distinct dimensions of support and undermining (Belsky et al. 1996; McHale 1995). Supportive coparenting is characterized by parents acknowledging each other’s competence and contributions, valuing each other’s involvement, and upholding each other’s authority. In contrast, undermining in the coparenting relationship refers to the extent to which parents criticize, blame, or disregard each other, and compete with one another for the child’s attention or loyalty (Teubert and Pinquart 2010).

The quality of coparenting has been linked to a host of family outcomes. For instance, coparenting alliances characterized by support have been associated with greater relationship satisfaction, better quality of parenting, and positive parenting behaviors, such as father involvement and maternal monitoring (Fagan and Lee 2011; Gavin et al. 2002; Jones et al. 2005; Schoppe et al. 2001). Additionally, an increasing number of studies have investigated the relationship between coparenting and children’s adjustment. For instance, the quality of the coparenting relationship has been negatively associated with child and adolescent externalizing and internalizing symptoms (Schoppe et al. 2001; Teubert and Pinquart 2010). Parents’ perceptions of supportive coparenting also have been linked to children’s academic competencies and social adaptation in school (Cabrera et al. 2012).

Although the role of supportive coparenting quality in the development of constructive parental alliances and positive family outcomes has been emphasized in recent decades, limitations in the literature exist. First, few studies examine stability and change in the quality of the coparenting alliance. Bonds and Gondoli (2007) found that among older, married mothers and fathers, marital adjustment was linked to improvement in coparenting quality over time. Similarly, Dush et al. (2011) found that greater romantic relationship commitment led to increases in supportive coparenting quality across several time points in a sample of African American and Hispanic parents. However, these studies have focused on older couples in married or previously committed relationships. Tracking the trajectory of coparenting quality in young, ethnically diverse parents who tend to be unmarried and have unstable romantic relationships presents an important line of further research. Second, there exists a dearth of longitudinal investigations into specific factors that might enhance the way parents work together to raise their children, particularly during the transition to parenthood.

Given the unique challenges that adolescent coparents often face, including relationship instability, financial dependence, and lack of parental competence (Beers and Hollo 2009; Hamilton et al. 2015), social support may be a particularly salient protective factor. In the current study, social support is defined as the presence of a strong relationship in which the other person shows care and affirmation, gives useful information and advice, and provides needed help and resources (Abbey et al. 1985). Two important sources of support for young parents that have been identified in the literature include their child’s other biological parent, or their coparenting “partner,” and the parents of young parents, or the “grandparents” of the child (Levitt et al. 1986; Unger and Wandersman 1988). Among young parents, social isolation and low support has been linked to depressive symptoms and parental stress (Barnet et al. 1996; Birkeland et al. 2005; Reid and Meadows-Oliver 2007), whereas social support has been linked to positive psychological adjustment and parenting (Barnet et al. 1996; Brown et al. 2012; Umaña‐Taylor et al. 2013). Of note, despite the importance of fathers’ involvement with their child and their child’s mother, and the challenges young fathers face in remaining involved, most research on family and partner social support on parent outcomes tend to exclusively examine mothers (e.g., Birkeland et al. 2005; Brown et al. 2012; Gee and Rhodes 2003; Taylor and Roberts 1995; Umaña‐Taylor et al. 2013).

In exploring the associations between partner social support and coparenting quality, researchers have generated mixed results. On the one hand, positive associations between partner support and coparenting have emerged. One study (Holland and McElwain 2013) found that positive marital quality was associated with mothers’ and fathers’ positive perceptions of coparenting quality. Similarly, Van Egeren (2003) suggested that the prenatal marital relationship, and especially fathers’ positive marital interactions, were important predictors of whether both parents reported experiencing a supportive coparenting relationship after the child’s birth. These results provide evidence that support in the marital relationship may cultivate support for each other as coparents.

Conversely, other previous research has demonstrated a lack of association between partner support and the quality of coparenting. For instance, one study found that within married, two-parent households, parents’ report of positive coparenting did not necessarily forecast positive relationship quality, and vice versa (Margolin et al. 2001). In investigating longitudinal associations between the marital relationship and coparenting, researchers discovered that positive aspects of the marital relationship at the 6-month time point did not predict supportive coparenting at the 3-year follow-up (Schoppe-Sullivan et al. 2004). However, this study focused on observations of a range of positive behaviors (e.g., engagement, cooperation, positive affect) in the marital relationship rather than primarily examining support between partners; additionally, results were based on a sample that largely consisted of low to middle income, Caucasian couples (Schoppe-Sullivan et al. 2004). Further examination of the association between support and coparenting quality among young, ethnically diverse parents is warranted, as the majority of previous studies have focused on older, married, Caucasian couples.

Notably, African American and Latino families tend to be more communal and interdependent, placing more value on interpersonal connections, compared to their Caucasian counterparts (Markus and Kitayama 1991). Adolescent parents commonly live in three-generation households, relying on their own parents for social support along with financial and material assistance and parenting help (Edwards et al. 2012; Jones Zalot et al. 2007; Oyserman et al. 1994; Roy et al. 2010). Grandparents who take on a variety of childrearing responsibilities are often regarded as non-traditional or non-marital coparents (Gonzalez et al. 2014; Jones et al. 2007). Both qualitative and quantitative research on African American families have demonstrated the significant role of extended family members, including grandparents, from contributing financial support to sharing childrearing responsibilities (Gonzalez et al. 2014; Taylor and Roberts 1995). Along similar lines, studies focusing on Latino families have demonstrated the focal role of elder relatives, particularly grandmothers, in supporting young parents through caregiving and other interfamilial processes (Burnette 1999; Umaña‐Taylor et al. 2013). The concept of familismo, which describes a strong orientation and dedication towards the family, is reflected in the tradition of intergenerational coparenting that is seen among Latino families. Therefore, examining the role of social support from grandparents may be particularly relevant for this high-risk group of young African American and Latino/a parents.

The presence of a supportive family network has been found to offset the potential difficulties associated with adolescent parenthood. In particular, social support from grandparents has been associated with positive couple/family outcomes. For instance, positive relationships between maternal grandmothers and fathers have been linked to more affectionate and supportive relationships between the individuals in the couple (Krishnakumar and Black 2003). Additionally, a supportive relationship between parents and maternal grandmothers has been associated with more consistent paternal involvement with the child over time (Cox and Bithoney 1995). These findings suggest that the presence of support from young parents’ own parental figures may be protective: grandparents may provide a supportive climate by encouraging and teaching their children to coparent more effectively.

Although there is some evidence to suggest that grandparent support may bolster the coparenting relationship, grandparent involvement may also hinder. Previously, grandparent support has been linked to poorer parenting outcomes, a finding that has been attributed to adolescents’ dependence on their parents and consequent ambivalence about the support due to their developmental desire to become more self-sufficient (Oyserman et al. 1994). In a similar way, when grandparents act as non-traditional or non-marital coparents, adolescent parents may be less motivated to improve their coparenting relationship with their child’s other biological parent. Thus, grandparents may play a gatekeeping role against sustained coparenting efforts between adolescent mothers and fathers. Furthermore, conflict regarding parenting roles and other child-related topics could lead to tension within both the grandparent-parent dyad (Bogat et al. 1998; Rhodes and Woods 1995) and in the mother-father dyad (Danziger and Radin 1990; Kalil et al. 2005). In line with family systems theory (Bowen 1976), triangulation, in which a person from a dyad (mother–father, parent–grandparent) seeks support from a third party individual, is more likely to occur during times of conflict. These relationship strains, in turn, could negatively impact the quality of coparenting.

Most coparenting interventions to date have focused on individual-level factors (e.g., emotion regulation, psychoeducation) and conflict resolution between parents (e.g., mediation) (Feinberg and Kan 2008; McHale et al. 2012; Teubert and Pinquart 2010). While attention to these topics is warranted, less is known about factors outside of the coparenting relationship that could enhance coparenting quality and be incorporated in such interventions. Ecological systems theory (Bronfenbrenner 1979) suggests that the partner relationship between parents is embedded within various interrelated systems that influence aspects of the coparenting relationship. For example, this theory implies that proximal processes within the immediate family system, such as interactions between mothers and fathers, are important to parent as well as child outcomes. By the same token, distal factors, such as social support from grandparents, can impact processes related to both the partner and coparenting relationships (Gee and Rhodes 2003; Unger and Wandersman 1988). Although many studies have highlighted the positive impact of grandparent support, a more nuanced investigation of this factor may help to elucidate the complexity of the role of grandparents during adolescents’ transition to parenthood.

The current study of adolescent mothers and their childrens’ fathers investigated whether grandparent support moderated the relationship between partner support and change in coparenting quality between 6 months postpartum (Time 1) and 12 months postpartum (Time 2). It was hypothesized that the association between partner support at Time 1 and change in coparenting quality across the two time points would differ at different levels of grandparent support. Specifically, it was expected that the effect would be stronger for those experiencing high levels of grandparent support compared to those experiencing low levels of grandparent support.

Method

Participants

Participants were 136 low-income adolescent mothers and their babies’ fathers (68 coparents), who were part of a larger study of young parents’ transition to parenthood. The majority of the mothers and fathers identified as African American (64.7% and 67.6%. respectively), with smaller percentages of Latino/a (29.4% and 26.5%, respectively) and biracial (5.9% and 5.9%, respectively) participants. The majority of Latino/a fathers and mothers were from Central America (83.9% and 78.3% respectively). Of the parents who identified as Central American, 40.7% were from El Salvador, 22.2% were from Guatemala, 13.0% were from Mexico, 3.7% were from Honduras, and 1.9% was from Nicaragua. The mean age of fathers was 20.54 (SD = 4.10; range = 15–41) and the mean age of mothers was 18.19 (SD = 1.27; range = 15–21) at Time 1. All of the mothers were primiparous.

Procedure

Participants were recruited from community sites located in a large mid-Atlantic metropolitan area, including high schools, clinics, and organizations providing social services to young parents. In order to participate, parents were required to be pregnant with their first child and in the third trimester at the beginning of the study. Both the mother and the father had to consent to participate in the study and identify as African American and/or Latino/a; however, recruited couples were not required to be romantically involved. An informed consent form signed by a parent or guardian and an assent form signed by the minor were collected from participants under the age of 18.

Trained graduate and undergraduate research assistants interviewed each parent separately at the participant’s home or a community site, according to participant preference. Interviews typically lasted about 1.5 h and were conducted in either English or Spanish. Each participant received $30 in compensation after the completion of an interview. Interviews were conducted during the third trimester and 6 months, 12 months, and 24 months after the baby’s birth. Data from the 6-month postpartum (Time 1) and the 12-month postpartum (Time 2) interviews were utilized in the current study.

Measures

Demographic information

At Time 1, participants were asked to provide information regarding age, race, educational attainment, employment status, and living arrangements. Residence with the other parent and residence with at least one grandparent were coded dichotomously, with 0 = not living with the person of interest and 1 = living with the person of interest.

Partner and grandparent support

At Time 1, social support between parents was measured using 11 items from the Social Support and Undermining Scale (SSUS; Abbey et al. 1985). Participants rated how much their coparent engages in socially supportive actions (e.g., “How much does your baby’s other parent provide you with encouragement and reassurance when you need it?”) as well as undermining, or discouraging, behaviors (e.g., “How much does your baby’s other parent make you feel unwanted?”) Participants rated items in the SSUS on a five-point likert scale (1 = Not at All to 5 = A Great Deal). Four items in the measure were reverse coded such that all items with higher scores indicated higher level of support. Items were averaged to form a composite score for partner support for mothers and fathers, respectively. Reliability analysis yielded a Cronbach’s alpha of .93, indicating excellent internal consistency.

Similarly, social support from the parents of participants (“grandparents”) was measured using the SSUS (Abbey et al. 1985) at Time 1. Participants responded three times, once for each parental figure of interest: the maternal grandmother, the maternal grandfather, and the paternal grandmother. Perceived support from the paternal grandfather was not assessed. A composite grandparent social support score was created by combining and averaging these three grandparent support scores. Reliability analysis yielded a Cronbach’s alpha of .86, indicating good internal consistency.

While the SSUS (Abbey et al. 1985) has been validated using a comparable sample of African American couples (e.g., Vinokur et al. 1996), it has not been validated among samples of adolescent parents or parents who identify as Latino/a. However, the Cronbach’s alphas for the current sample suggest that it is a valid measure of social support for this population.

Coparenting quality

At Time 1 and Time 2, coparenting quality was measured by 20 items from the Parenting Alliance Inventory (PAI; Abidin and Brunner 1995), a self-report questionnaire that measures the degree to which parents believe they have a sound working relationship with the child’s other parent. Sample questions include “When there is a problem with [our child], we work out a good solution together,” and “[My child’s other parent] and I communicate well about [our child].” Participants rated items in the PAI on a five-point likert scale (1 = Strongly Disagree to 5 = Strongly Agree). Items were averaged to form a composite score for coparenting quality for mothers and fathers, respectively, with higher scores indicated higher levels of supportive coparenting quality. Reliability analysis yielded Cronbach’s alphas of .95 and .96 for the measure at Time 1 and Time 2, respectively, indicating excellent internal consistency.

Rather than using simple difference scores, which have problems in terms of reliability, discriminant validity, and spurious correlations, a latent change score was used to index change in coparenting quality across the two time points (Peter et al. 1993). Parcels, or averages of individual items from the PAI, were created as indicators of the latent variable, change in coparenting quality (Coffman and MacCallum 2005). First, factor loadings for items on the PAI were generated. Then items were ordered from large to small loadings and individually assigned to one of three parcels, or factors that comprised coparenting quality at each time point. The same parcel assignments were used across time points.

Data Analyses

Data from the current study were analyzed with IBM SPSS Statistics 22, Release Version 22.0.0.0 (SPSS, Inc., 2013) and MPlus 7.1 (Muthén and Muthén 1998). The sample was selected for couples with complete data on the variables of interest at Time 1 and Time 2; consequently, missing data were not present within this sample. Data were centered prior to analyses and bivariate correlations among the measured variables were examined using SPSS. The two independent variables were combined as continuous variables in the moderator term (partner support × grandparent support).

MPlus was used to generate factor loadings and create parcels for the latent outcome variable, change in coparenting quality. This latent variable was indicated by two observed variables, coparenting quality at Time 1 and coparenting quality at Time 2, which were in turn defined by three parcels, as previously described. Prior to analyses, a path diagram of the hypothesized structural equation model (SEM) that included relationships between change in coparenting quality and the three predictor variables, partner support, grandparent support, and partner support × grandparent support, was developed. This model was estimated using MPlus and based on maximum likelihood estimation. To account for non-independence in the data, the complex option in MPlus, which clusters by couple, was used; this accounts for non-independence in standard errors (Muthén and Muthén 1998). The following goodness of fit statistics for this model were examined: the Chi Square statistic (χ²), the Comparativ1998 e Fit Index (CFI), the Tucker Lewis Index (TLI), and the Root Mean Square Error of Approximation (RMSEA). According to conventions regarding cutoff criteria for fit indices, the hypothesized model is considered a good fit to the data if the Chi Square statistic is non-significant, the CFI and TLI exceed .95, and the RMSEA is less than .06 (Schreiber et al. 2006).

To facilitate testing for moderation, the regression parameters from the model were used to estimate simple predicted slopes at levels of the moderator variable, grandparent support. A standard practice was used to interpret the interaction, examining it at two levels of the moderator variable (Aiken and West 1991): a high level of grandparent support was defined as one standard deviation above the mean, whereas a low level of grandparent support was defined as one standard deviation below the mean. After splitting the moderator variable into high and low levels, change in coparenting quality was calculated for different levels of partner support. Again, a high level of partner support and a low level of partner support was defined as one standard deviation above and one standard deviation below the mean, respectively.

Results

Means and percentages of mothers’ and fathers’ age, race, educational attainment, and employment status were generated (see Table 1). The following demographic information was reported at Time 1. Approximately half (54.4%) of the mothers and approximately two-thirds (64.7%) of the fathers reported that they had earned a high school diploma or equivalency. The majority (64.7%) of mothers and 41.2% of fathers reported that they were currently in school. Additionally, close to two-thirds (64.7%) of mothers reported that they were unemployed, whereas 27.9% of mothers reported working full-time. In contrast, 25.4% of fathers indicated that they were unemployed, whereas more than half (54.4%) of fathers indicated that they were employed full-time. The majority of mothers (78%) indicated that they were romantically involved with their baby’s father. Approximately one-third of parents (39%) reported that they lived with their child’s other biological parent and almost two-thirds (61%) lived with at least one grandparent.

Table 1 Demographic statistics
Full size table

Correlations between the variables of interest and these demographic variables were examined to determine potential covariates (see Table 2). At Time 1, unemployment was associated with higher levels of grandparent support (r = .21, p = .01). In addition, at Time 1, residence with the other parent was associated with lower levels of coparenting quality (r = −.26, p = .002). Therefore, employment status and residence with the other parent were included as covariates in the model.

Table 2 Descriptive statistics and bivariate correlations between study variables
Full size table

Means and standard deviations of each measured variable in the model as well as correlations between the measured variables were generated (see Table 2). Grandparent support (Time 1) and coparenting quality (Time 2) were not correlated with each other (r = .12, p > .05). All other variables of interest were significantly positively correlated with each other, in line with the hypotheses.

Structural equation modeling was used to test the hypothesis that the relationship between partner support (Time 1) and change in coparenting quality would be moderated by grandparent support (Time 1). Two models were tested: one in which employment status and residence with the other parent were entered as covariates and another without covariates. Because results across the two models indicated nearly identical parameters, the simpler model (without covariates) is reported here. Thus, the best-fitting model regressed partner support (Time 1), grandparent support (Time 1), and the moderator term, partner support (Time 1) × grandparent support (Time 1), on latent change in coparenting. All of the independent variables as well as the residuals were allowed to covary. Fit indices indicated a well-fitted model: χ² (19) = 20.45, p = .37, CFI = 0.999, TLI = .997, and RMSEA = 0.02. Given that the interaction term was significant (b = 0.31, p < .01), it is not appropriate to interpret findings concerning simple associations between the individual predictor variables and the outcome of change in coparenting quality (see Fig. 1). This analysis revealed that partner support was positively associated with changes in coparenting quality when grandparent support was high; however, that association became weaker and changed direction for lower levels of grandparent support (see Fig. 2).

Fig. 1
figure 1

Structural equation model illustrating the relationship between the independent variables, partner support, grandparent support, and the moderator term (partner support at × grandparent support), and the dependent variable, latent change in coparenting: χ² (19) = 20.45, p = .37, CFI = 0.999, TLI = .997, and RMSEA = 0.02. T1 = 6 months postpartum, T2 = 12 months postpartum. ***p < .001, **p < .01, *p < .05

Full size image
Fig. 2
figure 2

Interaction between partner support and grandparent support on change in coparenting quality

Full size image

Discussion

The current study explored the associations between social support from different members of parents’ social networks and changes in the quality of coparenting in order to identify potential protective factors for first-time adolescent parents African American and Latino/a, low-income during the transition to parenthood. Specifically, the present study examined partner support and grandparent support as individual predictors of change in coparenting quality, and tested whether grandparent support moderated the relationship between partner support and change in coaprenting quality.

The interaction of support from partners and grandparents was significant, and a closer examination of the effect yielded interesting findings. Specifically, among parents who reported receiving high levels of grandparent support, those who also received high levels of partner support experienced an increase in coparenting quality, while those who received low levels of partner support experienced a decrease in coparenting quality. Not surprisingly, these findings suggest that supportive relationships with both grandparents and partners may diminish some of the parenting risks associated with young parenthood. Supportive grandparents may encourage and teach their children how to effectively coparent, which could account for the increase in coparenting quality in the context of high partner support (Cooley and Unger 1991). On the contrary, some parents who receive high levels of support from their own parents may be less motivated to improve their relationship with their baby’s other biological parent. Moreover, the length of time that grandparents help and their level of involvement may be important to consider (Jones et al. 2007; Voight et al. 1998). Whereas assistance during acute stressors or transition periods may be helpful, long-term grandparent involvement could hinder the development of parenting skills, which could in turn negatively impact the coparenting relationship with the father. In line with the concept of grandparents as non-traditional or non-marital coparenting (Jones et al. 2007), this could explain the decrease in coparenting quality in the context of lower levels of partner support among parents with highly supportive grandparents.

Meanwhile, an opposite pattern of results was found for parents who reported receiving low grandparent support: among this subgroup of mothers and fathers, those with high levels of partner support experienced a decrease in coparenting quality, while those with low levels of partner support experienced an increase in coparenting quality. Results suggest that low levels of grandparent support may negatively impact the quality of coparenting between parents even when parents have a supportive relationship with one another. Reports of low levels of grandparent support may be an indication that parents are not receiving helpful coparenting instruction or advice from their own parents. Similarly, low grandparent support may imply disagreements between parents and grandparents regarding how to coparent. For instance, as joint stakeholders in the child’s upbringing, parents and grandparents may clash over one another’s roles and level of involvement (Bogat et al. 1998). More than half of this study’s sample was living with at least one parent at 6 months postpartum. Indeed, within the context of three-generation households, grandparents may engage in gatekeeping, or behaviors that restrict a parent’s presence and activities within the family (Allen and Hawkins 1999), which may present another source of social strain. Thus, the absence of a coparenting role model or the presence of conflict with grandparents about how to coordinate childrearing with the other parent could explain the decrease in coparenting quality in the context of high levels of partner support. Continued research is necessary to clarify the conditions under which the presence of grandparent involvement is perceived as beneficial versus harmful to the coparenting relationship.

Notably, in contrast to previous research, low levels of partner support were associated with an improvement in coparenting quality among parents with low levels of grandparent support. This somewhat counterintuitive outcome may be better understood in the context of family systems theory (Minuchin 1985). Family systems theorists state that when there is stress or tension in a family subsystem, such as the mother–father subsystem or the parent-grandparent subsystem, the individual in distress may recruit the assistance of a third party, such as another family member (Bowen 1976). Consistent with this idea, some parents may seek support and learn how to coparent from family members other than grandparents who they do not perceive to be sufficiently supportive. This explanation seems plausible given the extensive kinship networks that share in childrearing in both African American and Latino families (Taylor and Roberts 1995). Alternatively, these parents may learn to manage the coparenting relationship from new romantic partners who may have their own children. Research in older, married samples suggests that some parents may recruit new romantic partners (e.g., stepparents) as mediators to dispel coparenting tension with ex-spouses (Schrodt et al. 2006). In this way, these mothers and fathers may find encouragement and assistance from others in their social network, improving the quality of their coparenting relationship despite perceiving their coparents and their own parents as relatively poor sources of support.

Despite expanding scholarly thought regarding the construct of coparenting and growing efforts to improve coparenting among at-risk parents in recent decades, present-day research on families has dedicated few efforts to clarify the interplay between young parents’ social support and the quality of their coparenting relationship. Ecological systems theory (Bronfenbrenner 1979) asserts the importance of studying the interactions between proximal (mother–father interactions) and distal processes (the role of grandparents and other distant family members) in order to better understand coparenting. Furthermore, broadly speaking, the majority of extant studies on social support and coparenting have recruited older, Caucasian, married or committed couples. Thus, in considering study strengths, this short-term longitudinal study explores an understudied line of coparenting research as well as extends generalizability of findings to adolescent, low-income, ethnically diverse parents. Overall, results correspond with ecological systems theory, demonstrating the significant impact of interactions between parent and grandparent support on coparenting change.

In terms of methodological strengths, the current study utilized latent change scores to measure change in coparenting quality from 6 months postpartum to 12 months postpartum. A parceling technique described in Coffman and MacCallum (2005) was applied to facilitate the creation of a latent variable representing change across time points. This overall approach mitigates potential issues associated with using simple difference scores. Data analysis also accounted for non-independence in couples, which helps to mitigate covariance within members of the coparent-dyad.

Limitations and Future Research Directions

Limitations of this study merit consideration and present directions for future research. First, the current study utilized a relatively small sample of parents with a comparatively smaller representation of Latino mothers and fathers. Although the current sample size allowed for the application of structural equation modeling, the small sample of Latino coparent-dyads did not allow for a comparison between African American and Latino parents, nor a comparison among Latino subgroups. Compared to their African American counterparts, adolescent Latina mothers are more likely to be married to their babies’ fathers (Dickson 2001). In addition, research on Hispanic families has demonstrated ethnic differences regarding fathers’ role perceptions and involvement: for instance, Puerto Rican fathers have demonstrated low involvement in terms of direct childrearing, which may be attributed to beliefs about gender roles, whereas Mexican fathers may be more likely to engage in childcare, especially when children are young (Toth and Xu 1999). Consequently, investigations could benefit from recruitment of a larger minority sample, which would allow for the examination of within group differences. Similarly, in the future, it would be important to examine whether effects differ for mothers versus fathers, as previous studies have suggested gender differences in models predicting the quality of coparenting (e.g., Margolin et al. 2001).

Second, the present investigation of coparenting quality analyzed only two waves of data. Future inclusion of an additional time point could enhance understanding of coparenting changes during early parenthood, giving insight to trends in coparenting quality within this sample of young, ethnically diverse parents. In considering directions for future analyses, structural equation modeling of latent growth curves may be appropriate. This approach has been successfully applied to a similar type of longitudinal inquiry on coparenting quality in Dush et al. (2011).

Third, limitations exist regarding the measurement of support in this study. The current study’s use of an aggregate of social support from several grandparent sources (i.e., by averaging together reports from the maternal grandmother, maternal grandfather, and paternal grandmother) has limitations. Measuring perceived support from each grandparent separately could provide valuable insight, given that maternal grandmothers tend to play central roles within both African American and Latino families (Jones et al. 2007; Umaña‐Taylor et al. 2013), and particularly since the role of grandfathers has been understudied in this area.

Future research should examine different types of social support in relation to coparenting quality. For example, social support can be broken down into specific subcategories of support, including emotional support (e.g., care, understanding), tangible assistance (e.g., monetary or material aid), and guidance (e.g., advice or information on how to parent; Gee and Rhodes 2007). Similarly, given the multidimensionality of coparenting, it would be interesting and important to examine other domains of coparenting beyond support/undermining, such as division of labor, childrearing agreement, and children’s exposure to conflict (Feinberg 2003; Teubert and Pinquart 2010). Furthermore, family intervention research may benefit from assessing young parents’ social networks, which could help to identify other important individuals (e.g., extended family members) who may be important to include in interventions. Continued efforts to understand, enhance, and incorporate adolescent mothers’ and fathers’ sources of social support may be critical in order to ultimately improve how they work together as parents.