Predicting Coparenting Quality in Daily Life in Mothers and Fathers

Abstract

The purpose of the current study was to examine contextual (daily relationship quality, daily stressors, daily work hours), parent (daily negative emotions, gender), and child factors (daily child-induced parenting stress) as predictors of mothers’ and fathers’ perceptions of daily coparenting quality. Mothers and fathers from 174 families completed a 14-day diary study. Utilizing multilevel modeling, our results suggest that daily coparenting is multiply determined. Indeed, daily fluctuations in coparenting were predicted by similar daily fluctuations in couple relationship quality, parent negative mood, parenting stress, and father work hours. Moreover, as daily risk factors accumulated the quality of daily coparenting deteriorated further, suggesting that a buildup of stressors and daily difficulties may be particularly detrimental to parents’ abilities to cooperate with one another and coordinate their parenting together on a daily basis. Overall, our results suggest that there are many avenues through which we can intervene in the family system to improve the quality of daily coparenting.

Introduction

The coparenting relationship refers to the way that partners work together—such as supporting or undermining one another—in rearing their children (Feinberg, 2003). The coparenting relationship is at the center of the family system and many family interactions (Feinberg, 2003) and has therefore been linked to important family and child outcomes. For example, more supportive and positive coparenting relationships tend to predict greater marital quality, fewer child behavior problems, and more secure child attachment (Brown et al., 2010; Belsky & Hsieh, 1998; McHale & Rasmussen, 1998; Schoppe et al., 2001; Schoppe-Sullivan et al., 2004; Teubert & Pinquart, 2010). Therefore, elucidating influences on the coparenting relationship can better assist researchers and those who work with families as they attempt to enhance family and child well-being. In the present study, we examine contextual (relationship quality, daily stressors, daily work hours), parent (gender, negative mood), and child variables (child-induced parenting stress) as predictors of daily coparenting quality. Unlike earlier work, which has assessed coparenting using point-in-time measures (e.g., Bonds & Gondoli, 2007; Schoppe-Sullivan et al., 2004, 2008), we use an intensive longitudinal data (ILD) design (daily diaries) to examine coparenting on a daily basis. This type of a study is inherently designed to capture the small variations that may occur in perceptions of coparenting from day-to-day within individual parents. Such a study is necessary to better understand how coparenting reacts to daily perturbations and changes in the family system. Indeed, during some of the initial conceptualizations of coparenting in two-parent families, Gable, Belsky, and Crnic (1992) suggested that “it is the day-to-day functioning of the coparenting relationship that provides… one important mechanism by which poor marriages both directly and indirectly affect child development” (p. 284, emphasis added).

Although much of the research on coparenting has focused on the early parenting years, our knowledge continues to be limited with regard to variability in the coparenting relationship from day-to-day. Studies have found moderate rank-order stability in coparenting observations as well as parental reports of coparenting during infancy and toddlerhood (Davis et al., 2009; Favez et al., 2006; Gable et al., 1995; McHale & Rotman, 2007; Van Egeren, 2003, 2004) and into the preschool years (Feinberg et al., 2012; McHale & Rasmussen, 1998; McHale & Rotman, 2007; Schoppe-Sullivan et al., 2004). It is likely that, although there may be moderate rank-order stability over long time frames, coparenting quality varies within families over shorter periods of time as the family system experiences perturbations (such as stressors, interparental conflict, and so forth) until equilibrium is again established via feedback loops or small adaptations take place. For example, research on other family relationships, such as the couple relationship, has demonstrated significant variability in relationship feelings within individuals over time (Totenhagen, Butler, Curran, & Serido, 2015), and recent research using the same sample as the present study showed that mothers’ and fathers’ perceptions of coparenting also fluctuate on a daily basis (McDaniel, Teti, & Feinberg, 2017). Intensive data designs across days are needed to better assess coparenting within individuals and families and across time. Examining daily coparenting and its predictors will provide information about family or contextual characteristics that can be targeted to improve coparenting and ultimately parent and child outcomes.

Conceptual Framework

In general, frameworks for understanding influences on parenting and coparenting suggest that coparenting is multiply determined by contextual, parent, and child characteristics (Belsky, 1984; Feinberg, 2003; Minuchin, 1985). Indeed, coparenting is often conceptualized as being at the center of family functioning with internal and external sources of stress and support influencing the functioning of the coparenting subsystem (Feinberg, 2003; Minuchin, 1985). In the current paper, we sought to examine characteristics from each of these domains at the daily level, and we chose characteristics that have commonly been shown to be important in the literature (e.g., couple relationship quality, parenting stress, etc.). Additionally, according to family systems theory parents within families are interdependent, and therefore the functioning and experiences of one partner may crossover into experiences of the other partner (e.g., Minuchin, 1985). Although coparenting can take various forms of supportiveness, undermining, childrearing agreement, and more, we focus on coparenting more broadly—examining whether coparenting appears to function better overall depending on what is happening that day. We also view individuals and family systems as having a particular set or amount of resources (e.g., personal mental health, time, energy, etc.), and these resources are either depleted or enhanced by the various sources of contextual, parent, and child-related stress or support we will discuss in this paper; this perspective aligns well with Conservation of Resources (CoR; Hobfoll, 1989). If the various factors become sources of stress on a given day, the resources available in the family for parents to invest in high quality coparenting may become depleted, leading to coparenting interactions of lower quality overall (e.g., less support, less coordination, less investment in joint parenting, greater chances for disagreement, etc.).

Contextual Influences on Coparenting

Couple Relationship Quality.

Important contextual influences on coparenting quality include couple relationship quality, daily stressors, and daily work hours. Indeed, the functioning of the couple relationship is intimately connected to the functioning of the coparenting relationship, as both are subsystems within the family that directly involve the couple (Minuchin, 1985). In fact, couple relationship satisfaction is one of the most important predictors of coparenting quality. Pre-birth relationship satisfaction has been shown to set the tone for the coparenting relationship after birth (Le, McDaniel, Leavitt, & Feinberg, 2016; McHale, Kazali, et al., 2004; Schoppe-Sullivan et al., 2007; Van Egeren, 2004), and after birth it has been shown to be consistently linked to the quality of the coparenting relationship (Le et al., 2016; McHale, 1995, 1997; Schoppe-Sullivan et al., 2004) with those in more satisfied relationships showing both more supportive and less conflictual coparenting.

For example, reported relationship anxiety such as worries about not being loved (Belsky, Crnic, & Gable, 1995), observed hostility between partners (Katz & Gottman, 1996; Margolin et al., 2001), worse observed marital problem solving (Margolin et al., 2001), and lower observed positive engagement (Schoppe-Sullivan et al., 2004) have been linked to less supportive coparenting. Observed marital distress has also been linked to more competitive and less harmonious coparenting, as well as greater imbalances of involvement and warmth in coparenting interactions (McHale, 1995). Undermining coparenting has been associated with many of the same marital functioning indicators as mentioned before, including but not limited to observed marital hostility (Katz & Gottman, 1996), reported problems with feeling close to others (Belsky et al., 1995), worse reported marital quality (Le et al., 2016; McHale, 1997), low positive engagement (Schoppe-Sullivan et al., 2004), and defensiveness (Margolin et al., 2001). Research has also shown that couple relationship feelings (e.g., closeness, satisfaction, etc.) vary from day-to-day (Totenhagen et al., 2015), and we have shown in prior research on the same sample as the present study that mothers’ and fathers’ perceptions of coparenting co-vary with their daily feelings about their couple relationship (McDaniel et al., 2017).

Daily Stressors.

The presence of daily stressors may temporarily alter coparenting. This may occur because stress can make parenting and being emotionally available for one’s children and family more difficult as one’s resources become depleted (e.g., CoR; Hobfoll, 1989). For example, researchers have indicated that daily stressors and hassles (e.g., work stressors, arguments, etc.) are negatively associated with couple relationship quality (Harper et al., 2000; Lavee & Ben-Ari, 2007; Schulz et al., 2004; Story & Repetti, 2006; Totenhagen & Curran, 2011).

Work Hours.

Although work is necessary and at times a rewarding part of life, having to manage both work and family can produce work-family conflict and role strain (Fellows, Chiu, Hill, & Hawkins, 2016). Aligning with our theoretical framework, as work hours increase there is less time available (and at times cognitive or other resources as well) to be involved and engaged in parenting (e.g. Baxter, 2007; Pleck, 2010). In one sense, as work hours increase the chance for work-family conflict increases, and prior work has shown that work-family conflict is negatively associated with couple relationship quality (Fellows et al., 2016). Similarly, related work has found that marital interactions are rated more positively on days when individuals work less, have more energy, and have more leisure time (Doumas, Margolin, & John, 2003). Specific to coparenting, some work suggests that greater work hours are associated with poorer coparenting quality (Jia & Schoppe-Sullivan, 2011), although at times the association may not be as strong for mothers’ work hours (Schoppe-Sullivan et al., 2008). This may be due to the different gendered expectations that exist in families as mothers and fathers often experience diverging roles either from necessity, ideology, or choice (Cowan & Cowan, 1992). For example, although views have been shifting slowly in recent years, there is often still increased salience of the parenting role for mothers as compared with fathers (e.g., Hays, 1998). Such differing role expectations may lead to the conflict between work and family having different meanings for mothers and fathers on a daily basis, especially when it comes to engaging in childrearing.

Parental Influences on Coparenting

Parent Depressed or Negative Mood.

Individuals experiencing depressed mood likely find it more difficult to work together with their partner and to parent their child, as they feel drained, are less able to tolerate frustration, and are more likely to misinterpret relationship cues (e.g., Lovejoy et al., 2000). For example, researchers have found that more depressed parents show fewer positive interactions and more negative interactions in parent-child interactions; depressed parents are also less likely to structure family interactions into regular routines for their child such as around bedtime, naptime, and mealtime (Lyons-Ruth, Wolfe, Lyubchik, & Steingard, 2002; Paulson, Dauber, & Leiferman, 2006). Furthermore, some researchers have shown that depressive symptoms and distress related to parenting link with both less supportive and greater negative coparenting (Elliston, McHale, Talbot, Parmley, & Kuersten-Hogan, 2008; McDaniel & Teti, 2012; Tissot, Favez, Ghisletta, Frascarolo, & Despland, 2016).

Parent Gender.

At times, the gender of the parent is an important determinant of coparenting perceptions. For example, some have found that fathers report higher satisfaction with their coparenting relationship than mothers (Van Egeren, 2004). However, mother and father reports of coparenting are moderately to highly correlated during infancy (e.g., Feinberg et al., 2012; Schoppe-Sullivan et al., 2004; Van Egeren, 2004). Parenthood can be a gendered experience, as mothers and fathers often experience diverging roles (Cowan & Cowan, 1992; Hays, 1998). These differing roles, experiences, and levels of involvement with their child may lead mothers’ and fathers’ perceptions of the coparenting relationship to develop differently as well as may moderate the influences on coparenting.

Child Influences on Coparenting

Child-related Parenting Stress.

Children can sometimes be “a source of stress and strain in the marriage, a barrier to intimacy, and a cause of conflict” (Belsky, 1990, p. 172), especially when parents perceive their children as difficult. Although the relationship can sometimes be complex with the effects of the child’s temperament depending on other stressors that are present in the family system (e.g., Schoppe-Sullivan et al., 2007), some researchers have found that poorer coparenting quality is linked directly with more reactive and negatively emotional children (Davis et al., 2009). In parents with young children, stressors in regards to parenting are often common (Crnic & Low, 2002). Parenting stress can lead to feelings of depression and daily negative mood (Bolger et al., 1989; Leigh & Milgrom, 2008), a breakdown in family functioning (Cummings & Davies, 1994; Gelfand, Teti, & Radin Fox, 1992), reduced sensitivity and warmth (Crnic & Low, 2002), and less positive affect (Belsky, Woodworth, & Crnic 1996; Crnic, Gaze, & Hoffman, 2005). Additionally, parenting stress can be distinct from stress in other domains of life, such as the marital relationship (Belsky, 1984), and parenting stress may be more strongly tied to parenting behaviors than stress in other domains (Deater-Deckard, 1998). Finally, some research suggests that the associations between parenting stress and coparenting quality are similar for mothers and fathers (Solmeyer & Feinberg, 2011).

The Accumulation of Risk Factors

Many of these contextual, parent, and child factors are often interrelated. For example and as mentioned earlier in the introduction, parenting stress and deteriorations in family functioning, parent emotion and depression, and parenting quality often co-occur (e.g., Bolger et al., 1989; Crnic & Low, 2002; Cummings & Davies, 1994). As another example, parental depression affects individuals’ cognitive resources and interpretation of social cues which can also lead to greater risk for child dysregulation, poor parenting quality, and worse marital functioning (e.g., Elgar et al., 2007; Paulson et al., 2006; Whisman, Davila, & Goodman, 2011). Additionally, external stressors and daily hassles may lead parents to be more irritable and critical which can begin a circular cycle of negative parenting and child negative emotion (e.g., Webster-Stratton, 1990). As stressors and risk factors accumulate within individuals and families, the coordination between partners in parenting their child may suffer to a greater extent. Again, these impacts on coparenting quality can be explained in part by a depletion in the resources available to maintain a high quality coparenting relationship (CoR; Hobfoll, 1989; e.g., personal mental health, time, energy, etc.).

The Current Study

Although we have presented some initial evidence that coparenting quality fluctuates from day-to-day in our prior work on this same sample (McDaniel et al., 2017), we do not currently understand which factors most influence fluctuations in daily coparenting. The current study expands on the prior work of McDaniel et al. (2017) in which we developed and provided some initial validity for a measure of daily coparenting quality, the Daily Coparenting Scale (D-Cop). We utilize the D-Cop in a 14-day diary study of 174 heterosexual families (both mother and father reports) to examine coparenting quality on a daily basis. The purpose of the current study was to examine contextual (daily relationship quality, daily stressors, daily work hours), parent (daily negative mood, gender), and child factors (daily child-induced parenting stress) as predictors of mothers’ and fathers’ perceptions of daily coparenting quality. As called for by McDaniel et al. (2017), we hope to better illuminate the factors that influence the quality of daily coparenting; this work could then be used to inform prevention and intervention efforts and further strengthen family relationships by stabilizing the quality of coparenting. Based on the previously summarized literature, we make the following hypotheses. We also test for the potential for gender differences, as well as partner effects due to the interdependency inherent in family relationships.

H1. On days when parents experience worse couple relationship quality, parents will also feel worse about coparenting quality.

H2. On days when parents experience more stressors outside the home, parents will rate worse coparenting quality.

H3. On days when parents work more hours, parents will rate worse coparenting quality.

H4. On days when parents experience more negative emotions, parents will rate worse coparenting quality.

H5. On days when parents experience more parenting stress related to child difficulty, parents will rate worse coparenting quality.

H6. Finally, we hypothesize that these daily predictors will each account for additional variance in the experience of daily coparenting, such that those individuals experiencing a greater accumulation of these risk factors (contextual, parent, and child) will experience worse coparenting quality on those days.

Method

Procedure and Participants

Participants included both mothers and fathers from 174 heterosexual couples with a young child who were a part of the Daily Family Life Project. Participants were currently living together in the United States and had a child age 5 or younger (M = 2.88 years, SD = 1.33; 55% female). We recruited families through three primary sources: (1) a database of families across a Northeastern U.S. state who had expressed that they were willing to be contacted by researchers, (2) announcements on parenting websites and listservs, and (3) flyers in community buildings such as doctor offices. As the study was conducted entirely online, families were not required to live in the city or state in which the study took place. Families resided in the following U.S. regions: 52% Northeast, 17% West, 16% South, and 15% Midwest.

Participants’ relationship length ranged from 2 to 23 years, with 92% in a relationship of 5 years or longer (M = 9.99 years, SD = 4.07). Most were Caucasian (93% for mothers, 89% for fathers), married (95%), had a Bachelor’s degree or higher (76% of mothers, 68% of fathers), and were not currently attending school (80%); 57% had more than one child. In families with more than one child, mothers and fathers were asked to choose one child age 5 or younger which they would both report on during the study. On average, mothers were 31.52 years old (SD = 4.41; range 20 to 42), husbands were 33.31 (SD = 5.04; range 22 to 52), and yearly household income was approximately $74,000 (SD = $39,000; Median = $69,000), but ranged extensively from no income to $250,000 with 21% of families reporting they were on some form of federal aid (e.g., medical assistance, food stamps, etc.). Also, 68% of mothers and 91% of fathers currently worked for pay (mothers’ work hours, M = 31.46, SD = 14.09; fathers, M = 41.69, SD = 11.56).

Approval was obtained from the university IRB. Participants were assigned a unique ID number which they used each day they entered responses into our online survey. This ID number was able to link partners within families and participants across days. After study enrollment and informed consent, participants first completed a baseline online survey via a secure server. Our demographic characteristic variables came from this survey. Then, approximately two weeks after finishing their baseline survey (M = 17.87 days, SD = 9.38) participants completed 14 consecutive days of the Daily Coparenting Scale (D-Cop) and other daily measures before bed. To better ensure that parents completed surveys independently of one another, parents were sent unique survey links to separate email addresses; they were also asked to not share their ID numbers or responses with one another. There were 21 participants who dropped out or who did not complete any daily surveys, leaving us with a sample of 345 parents (174 women and 171 men from 174 families). Of those who completed at least one day of the daily surveys (94% of full sample), participants completed an average of 11.76 days (SD = 2.94 days), with 87% completing 10 or more days, for a total of 4058 person-days of data.

Daily Coparenting Measure

Daily Coparenting Scale.

The quality of daily coparenting was measured by mother and father reports across 14 days using the Daily Coparenting Scale (D-Cop; McDaniel et al., 2017). This scale includes 10 items that were created based on a careful review of the coparenting literature and that obtain a sampling of the range of possible coparenting-related feelings and behaviors that parents may encounter on a daily basis. Across the items, parents report on their daily experience of the solidarity of the parenting team, cooperation, support, endorsement, disagreement, undermining, and fairness in the division of childcare tasks. Parents selected the response to each item that best describes how they worked together as parents today on a 7-point Likert-type scale ranging from 1 (Strongly Disagree) to 7 (Strongly Agree). Example items include “We cooperated in parenting” and “We upheld each other’s rules and limits to the child.” Negatively worded items were reverse coded, and then all item responses were averaged for each day to produce an overall coparenting score for each day. A higher score indicates perceptions of higher quality coparenting on that day. In prior work using the same sample as the present study, researchers have shown the scale to have good internal consistency and reliability to assess within-person fluctuation in coparenting across days (McDaniel et al., 2017). In the current study, the scale also showed good reliability.

Contextual-Level Predictor Measures

Daily Relationship Quality.

On a daily basis, participants also rated how they felt that day about their relationship with their partner in terms of love, closeness, satisfaction, commitment, conflict, and ambivalence (Curran, McDaniel, Pollitt, & Totenhagen, 2015; Totenhagen, Serido, Curran, & Butler, 2012). Participants responded on a 7-point scale ranging from 1 (not very much or just a little) to 7 (very much or a lot). Example items include “Today, how satisfied were you with your relationship with your partner?” and “Today, how much conflict did you have with your partner?” Negative items were reversed scored; then items were averaged, with higher scores indicating feeling more positive about the relationship that day.

Daily Stressors.

Parents responded on a daily basis to seven items frequently used in the daily stressor literature asking whether they had experienced specific stressors or anything that most people would consider stressful in the last 24 hours in various contexts; these items included (1) argument or disagreement with anyone, (2) anything that could have argued or disagreed about, but decided to let it pass, (3) stressor at work, school, or volunteer setting, (4) stressor at home, (5) a happening to a close friend or relative that turned out to be stressful for you, (6) personal health, and (7) anything else that was stressful that was not previously mentioned (Almeida, Wethington, & Kessler, 2002). Participants also rated who this stressor was with (e.g., spouse/partner, child, etc.). We were interested in assessing stressors that were not related to family interactions. Therefore, to eliminate potential overlap with other factors (e.g., couple relationship quality, parenting stress, etc.), we only included stressors that were non-family related. These items were summed on each day to produce an overall number of non-family stressors experienced each day for each participant.

Daily Work Hours.

Parents were asked whether they worked for pay that day. If the parent responded yes, then the parent reported the number of hours worked. For the variable used in our analyses, we counted days not worked as having worked zero hours.

Parent-Level Predictor Measure

Daily Parent Negative Mood.

In order to broadly capture parents’ experiences of negative daily mood, parents responded to three items regarding how much time they had felt certain emotions that day on a 5-point scale ranging from 0 (none of the time) to 4 (all of the time). The items were anxious, angry or annoyed, and discouraged or sad, and items were selected and combined from the POMS-15 version of the Profile of Mood States (Cranford, Shrout, Iida, Rafaeli, Yip, & Bolger, 2006; McNair, Lorr, & Droppleman, 1992). We averaged the items to produce a negative mood score with higher scores indicating greater negative mood.

Child-Level Predictor Measure

Daily Child-Induced Parenting Stress.

Parents responded to three items dealing with child difficulty from the Parenting Stress Index (PSI; Abidin, 1990). We adapted the items such that parents were responding about their experiences with their child that day. The items included (1) “Today, I felt that my child was very moody and easily upset;” (2) “Today, my child did a few things which bothered me a great deal;” and (3) “Today, my child was very demanding.” Parents responded on a 5-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). We averaged the items on each day with higher scores indicating greater child-induced parenting stress on that day.

Reliability of Daily Measures

As daily measures are made up of various types of variation (e.g., between-person and within-person variation), we utilized the methods recommended by leading experts (e.g., Mogle, Almeida, & Stawski, 2015; Shrout & Lane, 2012) to calculate the reliability of our measures to assess between-person variation and within-person change. We report the between-person reliability estimates in Table 1 and the within-person reliability estimates in Table 2. Reliability estimates are not reported for daily stressors as this was a count measure of whether stressors occurred in various domains; in other words, we would not expect parents to report experiencing a stressor in other domains simply because they report a stressor in one domain—therefore, internal consistency was not expected between the stressor items. We also do not report reliability for daily work hours as this was measured via a single item. Overall, our measures showed excellent between-person reliability and moderate to high within-person reliability.

Table 1.

Between-person correlations and descriptive statistics for average daily coparenting and other average daily variables for mothers and fathers

	Coparenting (D-Cop)	Rel. quality	Stressors	Parent neg. mood	Parenting stress	Work hours
Coparenting (D-Cop)	.58***	.77***	−.07	−.34***	−.21*	−.08
Relationship quality	.79***	.56***	−.09	−.42***	−.10	−.17†
Stressors	−.14	−.04	.12	.36**	.18	.09
Parent negative mood	−.41***	−.34***	.44***	.31**	.39***	.08
Parenting stress	−.38***	−.20*	.15†	.38***	.60***	−.27**
Work hours	.07	.06	.00	−.26*	−.16†	−.13
Mothers
Mean	5.97	6.24	0.41	0.76	2.01	2.76
Std. Dev.	0.69	0.69	0.40	0.48	0.71	2.54
Between-person reliability	0.99	0.99	--	0.96	0.98	--
Fathers
Mean	5.96	6.18	0.35	0.62	1.83	5.16
Std. Dev.	0.69	0.74	0.35	0.46	0.64	2.36
Between-person reliability	0.99	0.99	--	0.97	0.98	--

Note:

^***p < .001,

^**p < .01,

^*p < .05,

^†p < .10

Mothers’ between-person correlations are displayed above the diagonal, while fathers are below the diagonal. Between-person correlations between mothers and fathers are bolded, italicized, and displayed on the diagonal.

Table 2.

Within-person correlations between daily coparenting and other daily variables in mothers and fathers

Daily Variables	Coparenting (D-Cop)	Rel. quality	Stressors	Parent neg. mood	Parenting stress	Work hours
Coparenting (D-Cop)	.36***	.46***	−.03	−.22***	−.14***	−.03
Relationship quality	.49***	.38***	−.06*	−.37***	−.08***	−.04*
Stressors	−.03	−.02	.25***	.24***	.03	.16***
Parent negative mood	−.21***	−.25***	.26***	.20***	.19***	.10***
Parenting stress	−.17***	−.14***	−.01	.12***	.36***	−.06*
Work hours	−.07**	−.03	.24***	.13***	−.12***	.37***
Mothers’ ICC	.54	.59	.22	.49	.41	.42
Fathers’ ICC	.58	.62	.22	.53	.41	.22
Mothers’ within-person reliability	.83	.86	--	.63	.82	--
Fathers’ within-person reliability	.82	.89	--	.62	.83	--

Note:

^***p < .001,

^**p < .01,

^*p < .05.

Significance levels were calculated in Mplus and accounted for the nested nature of the data. Mothers’ within-person correlations are displayed above the diagonal, while fathers are below the diagonal. Within-person correlations between mothers and fathers are bolded, italicized, and displayed on the diagonal. ICC = Intraclass correlation, or proportion of total variance in daily variable due to between-person variation.

Results

Descriptives and Correlations

On average across the 14 days, mothers and fathers reported feeling satisfied with coparenting and their couple relationship, and reported a low amount of stress, negative emotion, and parenting stress (see Table 1). Also, fathers reported working more hours on average than mothers. However, these between-person averages mask the amount of variability from day-to-day within individuals. We examined the intraclass correlation (ICC) for each daily variable (i.e., proportion of total daily variance due to between-person differences) and the between- and within-person correlations for all daily variables utilizing a multilevel structural equation model (MSEM) in MPlus (see McDaniel et al., 2017; Muthen, 1994). Mplus latently split the total variance in the daily variables into their between-person (e.g., individuals differ from one another in average levels) and within-person portions (e.g., individuals’ scores differ day-to-day from their own average level), and all daily variables were allowed to covary freely at the between-person and within-person levels. We then examined the standardized model results at each level. This model adjusts significance levels to account for the nesting in our daily data.

The ICCs ranged from .22 to .62 (see Table 2), indicating that 38% to 78% (e.g., 1 - .22 = .78) of the variance in our daily variables was tied to within-person differences. These ICC values show that there is variability at both the between-person and within-person levels in our daily variables and suggest that examining our variables and potential effects at both levels is important. In terms of correlations, at both the between-person level (see Table 1) and within-person level (see Table 2), our variables were often related in the ways we would expect. Individuals with greater relationship quality, lower child-induced parenting stress, and lower negative mood rated higher daily coparenting quality on average (between-person), although non-family stressors and work hours were not significantly associated with coparenting at the between-person level. We also found similar correlations at the within-person level, suggesting that on days when individuals fluctuate from their own average on a predictor (i.e., relationship quality, parenting stress, negative mood) they experience similar fluctuations in their perceptions of coparenting quality (within-person). Again, non-family stressors was not significantly associated with coparenting quality at the within-person level, although fluctuations in daily work hours were associated with fluctuations in daily coparenting for fathers.

Predicting Daily Coparenting Quality

To examine our hypotheses, we entered contextual, parent, and child characteristics as predictors of within- and between-person daily coparenting in a two-intercept multilevel model (MLM) for dyadic intensive longitudinal data (Bolger & Laurenceau, 2013). Utilizing this two-intercept MLM properly accounts for the nested nature of our data. We also created dummy codes for mothers and fathers (as recommended by Bolger & Laurenceau, 2013), which allows for estimates for each to be modeled simultaneously for all predictors. We also allowed for an autoregressive correlation in the residuals across days, as prior work has shown that not accounting for this can bias standard errors and significance tests (Bolger & Laurenceau, 2013).

Predictor variables that vary across days (i.e., relationship feelings, stressors, work hours, negative emotion, child-induced parenting stress) were decomposed into individuals’ average level of that variable across time (e.g., trait level relationship satisfaction; between-person effect) and individuals’ daily deviations from their average (e.g., state level relationship satisfaction; within-person effect). This was done by (1) grand mean centering the daily variable, (2) calculating the mean level in that variable across days within individuals (trait), and (3) subtracting the created trait variable from each individual’s daily scores on that variable (state). Thus, the trait portion measures between-person differences in that predictor; the state portion measures within-person fluctuations around their own average trait level in that variable; and the trait and state variable are uncorrelated. Of particular interest to us and our hypotheses are the state level associations, which if significant indicate, for example, that on days when one is less satisfied with the couple relationship one also feels worse about the coparenting relationship.

The simple MLM equations for predicting daily coparenting quality with daily time-varying contextual, parent, and child characteristics are presented here (partner effects are not written out here) for mothers, and these same equations would be estimated simultaneously for fathers. Partner effects for daily predictors were also included in the model.

$$\text{Level1:}Coparentin{g}_{ti}={\beta }_{0i}+{\beta }_{1i}Da{y}_{ti}+{\beta }_{2i}State\mathrm{Re}lationshipQualit{y}_{ti}+{\beta }_{3i}StateStressor{s}_{ti}+{\beta }_{4i}StateParentNeg.Moo{d}_{ti}+{\beta }_{5i}StateParentStres{s}_{ti}+{\beta }_{6i}StateWorkHour{s}_{ti}+e_{ti}$$

$$\text{Level2:}{\beta }_{0i}={\gamma }_{00}+{\gamma }_{01}Trait\mathrm{Re}lationshipQualit{y}_i+{\gamma }_{02}TraitStressor{s}_i+{\gamma }_{03}TraitParentNeg.Moo{d}_i+{\gamma }_{04}TraitParentStres{s}_i+{\gamma }_{05}TraitWorkHour{s}_i+{\gamma }_{\mathrm{06...}}DemographicControl{s}_i+{\mu }_{0i}$$

$${\beta }_{1i}={\gamma }_{10}+{\mu }_{1i}$$

$${\beta }_{2i}={\gamma }_{20}+{\mu }_{2i}$$

$${\beta }_{3i}={\gamma }_{30}+{\mu }_{3i}$$

$${\beta }_{4i}={\gamma }_{40}+{\mu }_{4i}$$

$${\beta }_{5i}={\gamma }_{50}+{\mu }_{5i}$$

$${\beta }_{6i}={\gamma }_{60}+{\mu }_{6i}$$

At Level 1, we have the equation describing the within-person relationship of daily coparenting quality (𝐶𝑜𝑝𝑎𝑟𝑒𝑛𝑡𝑖𝑛𝑔_𝑡𝑖) to the daily within-person predictor variables (e.g., state relationship feelings, state stressors, etc.). At level 2, we entered between-person predictors and controls—such as the trait portions (i.e., average levels) of the daily predictors and demographic controls—and between-person random effects. In our final model, we included between-person random effects for mothers’ and fathers’ intercepts (µ₀), daily relationship feelings (𝜇_2𝑖), and daily stressors (𝜇_3𝑖). The model could not converge if we estimated random effects for all daily variables, and we therefore removed those other random effects (i.e., 𝜇_1𝑖, 𝜇_4𝑖, 𝜇_5𝑖, 𝜇_6𝑖). Of primary interest to our hypotheses are the estimates for the state-level variables, which if significant represent that daily fluctuations in these predictors within individuals are associated with daily fluctuations in coparenting.

As we explained above, we were able to estimate the between-person random effects for mothers and fathers for the intercept, state relationship quality, and state stressors as well as their covariation. We report a few of the random effect estimates here. There was a significant amount of variation around the intercept for mothers (0.19, p < .0001) and fathers (0.16, p < .0001), as well as around the estimate for the effect of state relationship quality for mothers (0.08, p <.0001) and fathers (0.05, p < .001)—although not around the effect of state stressors for mothers (0.002, p = .42) and fathers (0.004, p = .25). Mothers’ and fathers’ intercepts covaried at level 2 (0.08, p < .0001), there was significant covariation between mother and father coparenting level 1 residuals on a given day (0.07, p < .0001), and there was also a significant autocorrelation in the level 1 residuals across days (AR1 = 0.14, p < .0001).

We report the unstandardized model estimates for our fixed effects for mothers and fathers in Table 3. Examining the fixed effects for the “state” daily predictors (within-person effects), we found significant effects for daily relationship quality, parent negative mood, and parenting stress, although not non-family stressors. These results indicate that on days when parents experience worse relationship quality, more negative emotions, and greater parenting stress—as compared with their usual level—they also feel that coparenting went more poorly than normal. However, perceptions of daily coparenting quality did not fluctuate with daily stressors outside of the home. Nevertheless, we found a significant effect for daily work hours for fathers only—on days when fathers worked more than usual they perceived coparenting to go more poorly than usual. We also found a few significant partner effects. On days when one’s partner felt better about their couple relationship, coparenting quality was better; additionally, on days when fathers worked more than usual mothers also perceived coparenting to be worse that day. For the most part, our results support our hypotheses (H1, H3, H4, H5, although not H2 which dealt with stressors outside the home). Finally, our results supported our hypothesis (H6) that an accumulation of risk factors would bode poorly for daily coparenting quality. In other words, the daily variables, except for daily stressors, showed unique and additive effects on fluctuations in daily coparenting. Therefore, if parents experienced negative fluctuations in multiple factors their coparenting would fluctuate more dramatically than if they experienced negative fluctuation in only one factor.

Table 3.

Multilevel model predicting parents’ daily feelings about coparenting

	Fathers		Mothers
Fixed effects	Estimate	Std Error	Estimate	Std Error
Intercept	5.89***	(0.14)	5.90***	(0.11)
Day	−0.0017	(0.003)	−0.0006	(0.004)
Baseline Control Variables
Family income	0.00002	(0.001)	0.0006	(0.001)
Not Caucasian	−0.10	(0.12)	−0.30*	(0.15)
Not college graduate	−0.06	(0.09)	0.02	(0.10)
Multiple children	0.09	(0.08)	0.11	(0.08)
Relationship length	−0.02*	(0.01)	−0.003	(0.01)
Marital status	−0.27	(0.20)	0.12	(0.19)
Parent age	0.01	(0.008)	−0.002	(0.01)
Child age	0.01	(0.03)	0.005	(0.03)
Employment status	0.09	(0.14)	0.06	(0.10)
Trait (between-person) portion of daily variables predicting average daily coparenting
Relationship quality	0.62***	(0.06)	0.62***	(0.06)
Non-family stressors	−0.05	(0.11)	0.08	(0.10)
Negative mood	−0.11	(0.09)	0.04	(0.10)
Parenting stress	−0.24***	(0.07)	−0.13†	(0.07)
Work hours	−0.02	(0.02)	0.005	(0.02)
Partner relationship quality	0.03	(0.06)	0.16**	(0.06)
Partner non-family stressors	−0.02	(0.09)	−0.08	(0.11)
Partner negative mood	0.05	(0.09)	−0.03	(0.09)
Partner parenting stress	0.009	(0.07)	−0.04	(0.07)
Partner work hours	0.03	(0.02)	−0.02	(0.02)
State (within-person) portion of daily variables predicting daily fluctuations in coparenting
Relationship quality	0.38***	(0.03)	0.46***	(0.04)
Non-family stressors	−0.007	(0.02)	0.01	(0.02)
Negative mood	−0.08**	(0.03)	−0.06*	(0.03)
Parenting stress	−0.08***	(0.02)	−0.06***	(0.02)
Work hours	−0.01***a	(0.003)	0.007 a	(0.005)
Partner relationship quality	0.10***	(0.02)	0.08**	(0.03)
Partner non-family stressors	−0.02	(0.02)	−0.002	(0.02)
Partner negative mood	0.02	(0.03)	−0.005	(0.03)
Partner parenting stress	−0.01	(0.02)	−0.004	(0.02)
Partner work hours	0.007 b	(0.005)	−0.02***b	(0.004)

Note:

^***p < .001,

^**p < .01,

^*p < .05,

^†p < .10.

Estimates for fathers and mothers modeled simultaneously. Day is centered on day 1. Control variables were coded as follows: Not Caucasian (0 = Caucasian, 1 = other race), Not college graduate (1 = college grad., 0 = less education than college grad.), Multiple children (1 = multiple children, 0 = only one child in family), marital status (1 = living together, not married, 0 = married), employment status (1 = work for pay, 0 = does not work for pay). Except for the above mentioned controls, all other variables were grand mean centered. Family income was in $1,000 units. Daily variables were split into trait (between-person) and state (within-person) portions. Significant differences between mother and father estimates were tested, and significant differences are represented with matching superscripts (e.g., a, b).

Discussion

We examined contextual (daily relationship quality, daily stressors, daily work hours), parent (daily negative mood, gender), and child factors (daily child-induced parenting stress) as predictors of daily coparenting quality. Extending earlier cross-sectional and macro-longitudinal work on coparenting, we used a daily diary design to examine coparenting on a daily basis. We chose to focus our work at the daily level in order to capture the small variations that occur in perceptions of coparenting on a day-to-day basis within individual parents and to better understand how coparenting reacts to daily perturbations and changes in the family system—all with the hope of identifying factors that can be targeted by interventions to improve coparenting and family outcomes. To our knowledge, this study is the first to examine predictors of coparenting at the daily level.

Our results confirm frameworks for understanding parenting and coparenting (Belsky, 1984; Feinberg, 2003)—i.e., coparenting is multiply determined by contextual, parent, and child characteristics. This confirmation of prior work is important, as this study now indicates that many of the processes examined or theorized by prior research at the cross-sectional or macro-longitudinal level may also function similarly at the daily level—a fact that was not known prior to the current study. Indeed, we found that fluctuations in daily coparenting were predicted by similar fluctuations in daily couple relationship quality, parent negative mood, parenting stress, and work hours.

These results speak to the complexity of family systems. In line with prior macro-longitudinal work and theory (e.g., Belsky, 1984; Le et al., 2016; Schoppe-Sullivan et al., 2004), we confirmed that when the couple relationship no longer acts as a support to the coparenting relationship on a particular day, coparenting quality may be diminished in both partners. Fluctuations in parents’ psychological states were associated with fluctuations in the quality of coparenting, as parents experiencing more depressed or negative mood likely feel drained, are less able to tolerate frustration, and are more likely to misinterpret relationship cues (e.g., Atkinson et al., 2000). Indeed, prior work has shown some linkages between coparenting quality and depressive symptoms (e.g., McDaniel & Teti, 2012; Tissot et al., 2016). Likewise, parents who feel stressed by parenting their child’s difficult behavior are less able to invest in high quality coparenting. Moreover, we found that daily risk factors predicted daily coparenting in an additive manner. This supports our hypothesis that an accumulation of risk factors would be particularly detrimental to daily coparenting quality. These results confirm a Conservation of Resources (CoR; Hobfoll, 1989) perspective and suggest that as resources (e.g., personal mental health, time, energy, etc.) are depleted by the occurrence of each additional risk factor (e.g., daily stressors, relationship problems, etc.) high-quality coparenting is increasingly difficult to maintain on that day.

Gender differences were often not found in the strength of the associations between our predictors and daily coparenting. This suggests that although mothers often are the primary caregivers and gatekeepers of parental involvement (e.g., Schoppe-Sullivan et al., 2015), fluctuations in parents’ individual or couple well-being on a daily basis appear to influence mothers’ and fathers’ perceptions of their relationship together as coparents in similar ways. Such similarities across gender in parental experiences could perhaps be used as a starting point from which to build common ground between stressed or struggling coparents. In a sense, although they may divide employment, housework, and child caregiving responsibilities in different ways parents may gain shared meaning and mutual understanding if they can validate one another’s experiences of how contextual, parent, and child stress influence them on a daily basis.

Although stressors outside the home were not associated with daily coparenting, another potential source of contextual stress—parent work hours—was found to be associated with daily coparenting quality. We also found gender differences in the association between daily work hours and daily perceptions of coparenting. Fathers’ work hours were associated with both mothers’ and fathers’ perceptions of daily coparenting quality, such that on days when fathers worked more than usual both mothers and fathers felt worse about their coparenting on that day. In the current sample, fathers worked on average more hours than mothers, although there was still sufficient range in both mothers and fathers. This effect may be due to gendered ideas about what it means to be a good father/mother and how this relates to working versus engagement in parenting or perhaps due to the different expectations that are placed on fathers or mothers (Cowan & Cowan, 1992; Hays, 1998). Future work should continue to examine how work and family demands interact to influence coparenting quality on a daily basis, as the number of two-parent households where both partners work has increased dramatically (Pew, 2015) leading to increasing salience of work-family conflict.

Our results suggest that there are many avenues through which we can intervene in the family system to improve the quality of daily coparenting. Indeed, teaching parents skills to reduce the salience or spillover of negative emotions or the stress of a difficult child into parenting could improve the quality of coparenting. Additionally, implementing programs at work to stop the spillover of work stress into the family domain could help, especially for careers where work hours fluctuate from day-to-day. It is important to note here that it is not always individuals who work more hours who experience worse coparenting (between-person effect); instead, regardless of overall level of work hours on average, on days when a parent, especially a father, works longer hours that parent—and likely his spouse or partner—will likely struggle more with maintaining high quality family relationships (within-person effect).

Moreover, the strongest predictor of daily coparenting was fluctuations in feelings about the couple relationship. The current work suggests that focusing on the quality of the couple relationship seems like a promising avenue for future interventions and research to improve coparenting relationships at the daily level. Indeed, family systems theory and prior cross-section and macro-longitudinal research all support our claim that the quality of the coparenting relationship is intricately and proximally connected to the functioning of the couple relationship (e.g., Le et al., 2016; McHale, 1995, 1997; Minuchin, 1985).

Limitations and Future Directions

Although the daily diary design of the current study allowed us to more fully examine coparenting relationships closer to the process level, we note several limitations. The majority of the sample was White and fairly educated, although we had great diversity in income and geographical location in the U.S. Our within-person design allowed us to examine how coparenting fluctuated on a daily basis in connection with within-person fluctuations in other daily variables, regardless of individual differences between participants. However, it is possible that the strength of the within-person associations between our daily variables could vary by ethnicity and socioeconomic status. Future work should expand on our initial examination of daily coparenting quality into more diverse contexts. Like all survey research, it is unclear how strongly parents’ daily self-reports relate to their actual coparenting behaviors across days. Daily diary designs reduce participant bias and retrospection as compared with measures that ask participants to recall occurrences across weeks or months (e.g., Bolger, Davis, & Rafaeli, 2003), which may help our results to be more accurate or at least lend another type of useful data in comparison with more macro-longitudinal or general survey designs. Additionally, as the data were all self-report, shared method bias is likely present, which could potentially inflate the observed associations; however, another way of looking at our data and analyses is that we controlled for the quality of daily mood (which is often thought to color one’s perceptions of all other aspects of one’s life and relationships) and still found these significant daily associations between our variables and coparenting. Moreover, the current results still merit additional study into the potential causal mechanisms and processes. In other words, the current paper is an initial examination of these daily processes and confirms that there are daily (not only cross-sectional or macro-longitudinal) associations between these characteristics and coparenting. We call for future work to continue examining the many potential pathways at the daily and micro-process level to further understand how coparenting and these other factors change and influence one another on a daily basis. For example, we often called our daily variables predictors, although it is likely that daily coparenting quality also influences these other daily variables (e.g., parent mood, stress, etc.). Finally, in the current study, we examined levels of daily coparenting quality broadly (i.e., higher vs. lower quality overall). Future work may wish to expand upon the McDaniel et al. (2017) D-Cop measure to develop measures of the specific domains of daily coparenting (e.g., supportiveness, undermining, agreement, solidarity, etc.), which could allow for more nuanced views of how various aspects of coparenting change and are connected with other factors on a daily basis.

Conclusion

In the current study, we examined coparenting quality on a daily basis and found that its fluctuations are linked with fluctuations in parents’ everyday experiences. We confirmed that coparenting is multiply determined by a range of contextual, parent, and child characteristics. Indeed, we found that fluctuations in daily coparenting were predicted by fluctuations in daily couple relationship quality, parent negative mood, parenting stress, and father work hours. Moreover, as daily risk factors accumulated the quality of daily coparenting deteriorated further, suggesting that a buildup of stressors and daily difficulties may be particularly detrimental to parents’ abilities to cooperate with one another and coordinate their parenting together on a daily basis. Overall, our results suggest that there are many avenues through which we can intervene in the family system to improve the quality of daily coparenting—although coparenting may benefit most from interventions targeting the quality of the couple relationship. Parents who find ways to support one another as a couple (outside of parenting together) are the most likely to have high functioning coparenting relationships on a daily basis.