Is It OK If Mommy Dates? The Influence of Mother’s Post-Divorce Dating Transitions for Children’s Behaviors

Abstract

Despite many mothers’ interest in establishing romantic relationships following divorce, some worry about the impacts of dating transitions, such as starting or ending a dating relationship, for their children’s adjustment. The literature illustrates that transitions in and out of relationships could hinder children’s post-divorce adjustment, but a supportive relationship between children and mothers could buffer the damaging effects. Prior research mostly focused on repartnering (i.e., cohabitation and remarriage) whereas a comparative form of romantic experience, post-divorce dating, is understudied. Post-divorce dating, a critical precondition for cohabitation and remarriage, is assumed to have the power to shape family stability and functioning. The goal of this study is to examine the impact of mothers’ post-divorce dating transitions for children’s adjustment. Additionally, we test child-mother rapport as a moderator for the relationship between dating transitions and children’s adjustment. Longitudinal data for this study comes from mothers of elementary-aged children who were within three months of filing for divorce (N = 316). Results from hierarchical linear models show that mothers’ dating transitions exert little direct effect on children’s adjustment. However, child-mother rapport moderated the relationships between the time in mothers’ dating relationships and children’s prosocial behaviors, and between mothers’ romantic breakups and children’s externalizing behaviors. This adds further support that high child-mother rapport benefits children’s adjustment within the context of divorce. Overall, the results suggest that the post-divorce dating relationship transitions may influence children’s well-being via indirect pathways and the moderation effect of child-mother rapport on children’s behavior are inconsistent across dating transitions. Implications for family adjustment and future research are discussed.

Introduction

Divorce is a common experience in the United States, as divorce rates for first marriage have fluctuated between 40–50% over the past two decades, and the divorce rates for second marriages hover around 60% (Centers for Disease Control and Prevention, 2017; National Marriage Project, 2019). Divorce has been perceived as stressful and distressing for families (Amato, 2000; 2010; 2014), yet not all parents and children are impacted by divorce in the same way or by the same degree (Kelly & Emery, 2003; Lansford, 2009). One typical approach divorced parents take to cope with post-divorce stress is forming new romantic relationships, such as dating and committed relationships. Although divorced fathers date more often than divorced mothers (Amato, 2010), Langlais et al., (2016b) found that around 40% of American mothers formed multiple romantic relationships within the first two years following divorce. Indeed, the challenge of striking a balance between dating and parenting could be a barrier for single parents to pursue and establish a romantic relationship, and this barrier is more pronounced for women who typically dedicate more custodial time and energy to their children compared to men (Gray et al., 2016). Yet, the desire to meet one’s romantic needs as well as the potential emotional and financial support afforded by a romantic relationship still drives the divorced population to engage in romantic experience after divorce (Cartwright, 2010; Osborne et al., 2012).

Although research concerning post-divorce romantic relationship formation has advanced the understanding of post-divorce adjustment, these studies mostly focus on remarriage and cohabitation (Amato, 2010; Ganong & Coleman, 2000), which we refer to as repartnering. Specifically, we define repartnering as remarrying or moving in with a romantic partner, which are significant relationship events characterized by high levels of commitment. Understandably, repartnering relationships are hardly achieved overnight. Those who experience divorce, in many circumstances, will undergo a series of post-divorce dating transitions. In the current study, we define post-divorce transitions as initiating a dating relationship, maintaining that relationship, ending that relationship, and beginning a subsequent relationship (Langlais et al., 2016b). These transitions often precede and prepare for subsequent cohabitation and remarriage, presuming divorced parents enter repartnering. Besides functioning as a precursor of repartnering, post-divorce dating distinguishes itself from repartnering in a few notable ways. First, the seriousness of post-divorce dating relationships is generally incomparable to that of remarried and cohabiting relationships (Langlais et al., 2016a; 2016b). The latter types often reflect a higher relationship quality as the decision to move in and share a home together necessitates a strong relational commitment. Plus, the nature of interactions in dating relationships varies from that in remarriage and cohabitation. Indicated by Anderson and Greene (2005), post-divorce dating relationships consist of largely dyadic activities between the parent and the partner, whereas triadic activities between the parent, the child, and the partner emerge and become more frequent once the relationship develops into a more advanced stage represented by repartnering. Despite these differences, the vast majority of research on post-divorce family adjustment has focused on repartnering, leaving how post-divorce dating interplays with family understudied. To fill some of these gaps, the primary goal of this study is to explore the effects of mothers’ post-divorce dating transitions for children’s adjustment, which we measure using children’s externalizing, internalizing, and prosocial (i.e., helpful and supportive) behaviors. Divorced mothers are the focus of this study, as the stress experienced by this demographic typically exceeds divorced fathers, divorced non-parents, and single mothers (Hetherington & Kelly, 2003; Symoens, et al., 2014). After divorce, mothers often experience increased feelings of depression, loneliness, and anxiety, while also reporting significant declines in income and social network support (Hetherington & Kelly, 2003; Langlais et al., 2016a; Symoens et al., 2014).

Post-divorce Dating and Children’s Adjustment

The present study is primarily guided by the divorce-stress-adaptation perspective (DSAP; Amato, 2000) and emotional security theory (Davies & Cummings, 1994). Amato (2000) held that a divorce is not a one-time event but a process, which can bring about a multitude of stressors, termed as “mediators,” for both parents (e.g., decreased economic resources, emotional support) and children (e.g., less contact with one parent and continued exposure to inter-parental conflict). These stressors impose direct effects on the family’s adjustment, including poorer psychological health and more emotional and behavioral problems. Although DSAP makes no direct indication of post-dating transitions as a stressor, we make a case that post-divorce dating transitions could operate as impactful events that shape children’s adjustment.

One way to manage post-divorce stress for divorced individuals is to date and/or repartner. In addition to the inherent human desire to develop enduring romantic relationship across the life span (Fisher, 2016; Gray & Garcia, 2013), the formation of post-divorce repartnering relationships is associated with easing post-divorce adjustment for mothers (Hetherington, 2003; Langlais et al., 2016a; Hughes, 2000) and children (Kelly & Emery, 2003; Langlais et al., 2018). Some studies have found benefits for parents and children when dating partners contribute to the family, such as providing financial resources, assisting with housework, or helping children with academic or athletic obligations (Bzostek et al., 2012; Hetherington & Kelly, 2003). Langlais et al., (2018) found that when rapport between mothers’ dating partners and children was high, children were less likely to exhibit problem behaviors.

It is true that a nurturing post-divorce romantic relationship can facilitate children’s post-divorce adjustment; yet, family transitions during and after divorce (e.g., cohabitation, entering and exiting a dating relationship) are perceived as potentially deleterious to children’s psychological well-being and social competency (Bachman et al., 2012; Bzostek & Beck, 2011; Cooper et al., 2011). The effect of a new transition on children’s physical and emotional health may accumulate on top of the preceding transitions (Bachman et al., 2012). This implies children will likely have increasing difficulties managing their behavior and preserving their well-being as transitions are mounting up. Frequent parental relationship transitions could put children’s well-being at stake and heighten the risk for children to develop depressive symptoms, lower self-esteem, increased externalizing behaviors, and more parent-child conflict (Capaldi & Patterson, 1991; Coleman et al., 2000; Lansford, 2009). Meanwhile, multiple transitions test parent’s capability to navigate through a relatively unstable romantic life and could distract them from other responsibilities such as parenting. These issues justify single mothers’ prudence in making decisions around their dating. A qualitative study by Hadfield and Nixon (2017) investigated single mothers’ perceived benefits and barriers when attempting to form a romantic relationship. The results revealed that some mothers refrained from building a relationship because of the concerns over curtailed time and attention devoted to their children. Moreover, these mothers were anxious about the changes in paternal involvement with the child and quality of co-parenting when the non-residential fathers found out the mother was dating.

Despite the dating dilemma confronted by single parents and children’s tendency to exhibit more problem behaviors when a family goes through numerous post-divorce transitions, extant literature implies the possibility of changes in prosocial behaviors when certain conditions are met. Through the lens of emotional security theory (Davies & Cummings, 1994), a parent’s entering, staying in, or terminating a dating relationship could lead to changes in children’s prosocial behavior as these transitions influence children’s sense of emotional and attachment security. If transitions are viewed as reasonable and beneficial rather than stressful, and the emotional bond between children and mothers are secure, mothers’ dating experiences may invoke more socially desirable behaviors in children instead of interfering with children’s adjustment. For example, children having attachment security were found to engage in more prosocial behaviors (Paulus et al., 2016) and less aggressive behaviors (Deborah, 2006) during stressful experiences. A longitudinal study by McCoy et al. (2009) illustrated a positive association between children’s prosocial behavior and their sense of emotional security toward their parents’ marital relationship.

The literature presents mixed findings. Studies suggests that the adjustment of children may be undermined by constant family transitions following divorce while some children fare well and exhibit prosocial behavior as a result of a sustained supportive family environment. In parallel, existing as a distinct form of relational status from repartnering, parents’ post-divorce dating could play a crucial and independent role in shaping family stability and functioning (Zito & De Coster, 2016). Given the gaps and limited understanding, it is worth investigating the effect of mothers’ dating transitions on children’s adjustment, which, in the present study, is operationalized as externalizing, internalizing, and prosocial behavior. Since theories and empirical evidence do not point out the directionality regarding the impact that these transitions may have for children’s behaviors, we proffer the following hypothesis:

Hypothesis 1

Mothers’ post-divorce dating transitions (initiating, maintaining, ending, and forming a subsequent dating relationship) will be associated with children’s internalizing, externalizing behaviors, and prosocial behaviors.

The Moderation of Child-Mother Rapport

Another objective of this study is to explore to what extent rapport between children and mothers moderates the relationship between post-divorce dating transitions and children’s behavior. Within the DSAP (Amato, 2000), moderators constitute a critical group of variables influencing how well individuals cope with post-divorce stressors. Intra-individual factors (e.g., self-efficacy, social competency), inter-personal relationships (e.g., support from family members and peers), and components in the broader context (e.g., community service) can serve as buffers attenuating harmful effects of stressors on family well-being. Amato (2000) did not directly identify child-mother rapport as a moderator, but it is theoretically and empirically reasonable to predict that this relationship has the ability to modulate the impact of post-divorce dating transitions for children’s adjustment given that the mother-child bond is one of the most central interpersonal relationships for children’s development.

The child-mother relationship plays a vital role in children’s adjustment (Jensen et al., 2018), especially during times of family change (Mooney et al., 2009). For instance, Eldar-Avidan and colleagues (2009) interviewed 22 Israeli emerging adults who were asked to reflect on their perceptions, beliefs, and feelings they held toward their parents’ divorce, which occurred when these participants were between nine and ten years old. The results classified participants into three profiles—resilience, survival, and vulnerability. The defining distinction between resilience and the other two profiles is the persistent presence of a supportive relationship with the custodial parent. The subjects coded as resilient regarded the relational bond with parents, particularly mothers, as the most prominent source of resilience and strength. Additionally, studying first graders whose parents divorced, Bezuidenhout et al. (2018) found that, in addition to intrinsic factors (i.e., sense of agency and inner motivation), children’s supportive social system—their relationship with parents, siblings, and extend family members—promoted resilience to tackle both familial and academic challenges. In a study by Dillman Taylor et al. (2011), divorced parents attended a 10-week child-parent relationship therapy program, the aim of which focused on enhancing child-parent interactions by teaching parents child-centered skills and principles. Parents reported improvement in their child’s externalizing and internalizing behaviors as a result of this program. Stadelmann et al. (2010) posited that the child-parent relationship plays a moderating role between parental separation and the development of children’s behavior problems. Existing studies affirm that high quality child-mother relationships can help protect children from the damaging effects of family-related stressful events. In the present study, we employ child-mother rapport as the proxy measure of quality of child-mother relationship. Taken together, we proffer the following hypothesis:

Hypothesis 2

Child-mother rapport will moderate the relationship between mothers’ post-divorce dating experiences (i.e., initiating, maintaining, ending, and forming a subsequent dating relationship) and children’s behaviors.

Method

Participants

Data for this study comes from a multi-method, longitudinal study of mothers’ romantic relationships and family adjustment after divorce (Anderson & Greene, 2005). Within 60 days of filing for divorce, mothers were recruited to participate in this study. Eligibility included having at least one elementary-school aged child (5 to 11 years old) who resided with the mother at least 50% of each week (N = 316). Of the mothers who participated in this study, 77% were experiencing their first divorce, 15% their second divorce, and 8% of mothers were experiencing at least their third divorce. The average length of marriage was 122.26 months (Range: 8 months–321 months; SD = 65.21) and the average length of separation from the former spouse was 14.56 months (Range: 0–103 months; SD = 20.62). Children who participated in the study (only one child was selected per family at random) were the biological or adoptive children of the parents who were ending the current marriage. Fifty-two percent of the children were female and the mean number of children living in each household was 2.07 (Range: 1–6; SD = 0.90). The average age of children was 7.77 (Range: 4–12; SD = 1.97) and the mean age of mothers was 36.8 years old (Range: 21–53; SD = 6.59). Sixty-four percent of mothers identified themselves as white, 27% as Hispanic, and 9% as African American. Mothers’ education levels are as follows: 8th grade or less (1.9%), some high school, not graduated (2.8%), specialized technical degree, but did not graduate high school (0.9%), GED (3.8%), high school graduate or equivalency (not GED; 8.8%), specialized technical degree after completing high school (6.6%), two years or less of college (19.7%), Associates degree (5.6%), more than two years of college, but no degree (12.2%), Bachelor’s degree (23.2%), one or more post-graduate years of college, but no post-graduate degree (5.3%), Master’s degree (7.8%), and advanced doctoral degree (i.e., PhD, MD; 1.3%). Although 82% of mothers were working in a paid position at least part-time, 23% of mothers received some means of governmental assistance.

Procedures

The current study was approved by the appropriate institutional review board. Mothers provided their own informed consent at the beginning of each assessment. For assessments that involved children, mothers provided parental consent and children provided their own assent prior to participating. Divorce petitions were obtained from public court records in a large metropolitan city in the south-western United States. Any mother who (or whose spouse) filed for divorce within a period of 60 days during the spring of 2004 (until spring 2006) were mailed recruitment brochures to the most recent address provided. Brochures were mailed to 987 mothers. The brochure explained the purpose of the study and informed mothers about a follow-up telephone call that would describe the study, answer any questions they had, and verify eligibility for the study. Each brochure contained $1.00 as incentive for mothers to read its contents and respond to the follow-up telephone call, which was used to schedule a get-acquainted visit, where the research team would come to the participants’ home and elaborate on details of the study. The phone calls were conducted by an independent research firm, and data regarding the number of calls made and answered were not collected. As a result of these calls, 363 eligible families agreed to participate in a get-acquainted visit in the family’s home to discuss the study in greater detail. Eighty-eight percent of families who agreed to the get-acquainted visit accepted participation in the study (N = 316). A few weeks later, these families participated in an initial assessment, which we refer to as the “baseline assessment,” where mothers answered demographic questions and questions concerning their family and divorce status. Before the initial assessment, the research team chose a target child within the specific range using stratified random sampling to have equal numbers of male and female child participants in the current study. More specifically, mothers interested in the study provided ages and gender for each of their children during the “get-acquainted” visit; children in the specific age range required for this study (i.e., elementary aged, 5–12) were separated according to gender, and randomly selected within these groups to ensure similar number of male and female children.

Although this study involved multiple assessments, data for this study comes from monthly surveys that mothers completed over the course of 24 months starting at the baseline assessment. Mothers were given the option of completing the monthly surveys online through a password-protected website or through the mail. Mothers provided information for one or multiple romantic relationships per survey (if they were dating), including the date a relationship began or ended and their levels of satisfaction and commitment with each romantic partner. Mothers also answered questions concerning their children’s externalizing, internalizing, and prosocial behaviors. The mean number of monthly surveys that mothers completed was 14.58 (out of 24 possible; Median = 13; SD = 9.71; Range: 1–24), and 91% of mothers completed at least three surveys. Mothers were compensated $5 for each monthly survey. Loss of participants over time occurred due to various factors, including: ineligibility (primarily reconciliation or changes in custody; 5%), temporary loss of contact (e.g., participants completed the 11-month and 13-month assessment but not the 12-month assessment; 10.7%), and dropout (6%). Independent samples t-tests were conducted to examine mean differences among participants who participated in the study and those that dropped out regarding the independent and dependent study variables and no significant differences were found.

Eligibility was not based on whether a mother had started dating or not and there was not a requirement for the child to know whether their mother was dating. Accordingly, at the baseline assessment of the study, 44.5% mothers were in a new relationship after filing for divorce, 26.5% reported interest in dating, and 29% reported no interest in dating. At the end of the study (two years after the baseline assessment), 86% of mothers stated that they were in a serious dating relationship and 24% of mothers reported at least one breakup of a serious relationship at some point during the study. More precisely, 160 mothers reported a total of 205 dating breakups: 119 mothers reported one breakup, 37 mothers reported two breakups, and four mothers reported three breakups. During the study, 27 mothers remarried and 109 mothers entered cohabiting relationships.

Prior to analyses, we examined mean differences for children’s behaviors between mothers who dated during the study (N = 270) versus mothers who did not date during the study (N = 46), as well as differences in maternal characteristics. Although there were no differences in externalizing and internalizing behaviors, children of mothers who did not date reported more prosocial behaviors compared to children of mothers’ who did date (F(1,315) = 9.94, p < 0.01). There were no significant demographic differences between mothers who did and did not date.

Measures

Post-divorce dating initiation and length of relationship

Mothers reported their romantic relationship status on their monthly survey for each person they dated over the last month. Mothers selected one of the following options for each potential or current partner: 1 (Interested but not yet romantically involved), 2 (Never romantically involved, and now no longer interested), 3 (In a casual romantic relationship), 4 (In a serious romantic relationship), and 5 (Romantic relationship was casual or serious, but has ended). A discrete-time, person-period dataset was created, with each line corresponding to a monthly survey a mother completed. In this dataset, a dichotomous variable signaled whether a mother began a casual or serious dating relationship for that month. A reporting of 1 or 2 by mothers was coded as 0 (not dating) and a reporting of 3 or 4 was coded as 1 (dating). Another variable was created to measure the time spent in each relationship (time since first reported a 3 or 4 to the question above) until a breakup or the end of the study. These two variables capture the intercept (the point in which mothers enter a serious dating relationship) and slope (time in that relationship) of mothers’ first post-divorce dating relationship during the study. During the time when a mother was not dating, the intercept and slope variables reset back to zero.

Maternal breakup

Mothers who broke up with a romantic partner responded with a 5 to the romantic relationship status question aforementioned. To determine breakup for the current study, times when mothers reported a 5 after it followed a 3 or 4 (casual or serious dating relationship) were included as breakup. The date that the breakup occurred triggered a dichotomous variable to alternate from a zero to one (0 = not experienced a breakup; 1 = experienced a breakup), which represents the intercept when a mother first reports a post-divorce breakup. A slope variable corresponding to time since breakup was recorded until the mother entered a new relationship or reached the end of the study.

Mothers’ relationship quality

Mothers responded to the following question concerning satisfaction with her dating relationship on her monthly survey: “All things considered, how happy or unhappy has the relationship with this person been this past month?” Response choices ranged from 1 (Very Happy) to 6 (Very Unhappy) and was reverse-scored for ease of interpretation. Mothers also reported their romantic commitment by responding to the item, “How likely is it that you will have a long-lasting or permanent, romantic relationship with this person?” on their monthly survey with response choices ranging from 1 (Very Likely) to 5 (Very Unlikely), which was also reverse-scored. Mothers reported an average score of 3.35 (SD = 2.52) for satisfaction and 2.63 (SD = 1.58) for commitment per monthly survey. Because these measures were highly correlated, r (4228) = 0.94, p < 0.01, they were summed to provide a single measure of relationship quality (Range: 0 – 11; M = 5.98; SD = 4.10).

Perceived children’s behaviors

To assess children’s behavior, we employed Behavior Problem Index (Zill & Peterson, 1986). On each monthly survey, mothers answered 20 items that asked if children exhibited externalizing (6 items), internalizing (7 items), or prosocial (7 items) behaviors. Example items included, “In the PAST 24 h has your child helped someone?” (prosocial behavior), “In the PAST 24 h has your child had a fit or temper tantrum?” (externalizing behavior), and “In the PAST 24 h has your child been depressed or sad?” (internalizing behavior). Mother reported either Yes (1) or No (0) for each item. Given this response scale, we examined the data to ensure variability within mothers across all monthly surveys, where 9.2% of mothers reported no variability in children’s externalizing behaviors, 11.1% reported no variability in children’s internalizing behaviors, and 10.4% reported no variability in prosocial behaviors across the monthly surveys (of these, 4.7% of mothers reported no variability across all children’s behaviors, and 2.2% reported no internalizing and externalizing behaviors during the course of the study; we included these mothers in the current sample, as it is possible that these children may not have experienced variability in their behaviors over the course of the study, but note later the potential of maternal bias towards their children’s behaviors). Individual item means for children’s behavior was 0.22 (SD = 0.25) for externalizing; 0.15 (SD = 0.20) for internalizing; and 0.85 (SD = 0.18) for prosocial behaviors across all time points in the study. The scales demonstrated adequate internal consistency across the study (Cronbach’s alpha = 0.76; 0.75; 0.71 respectively).

Child-mother rapport

In addition to the 24 monthly surveys, mothers also completed larger questionnaires in their homes at the initial assessment, and 12- and 24-months after the initial assessment. At these assessments, mothers reported their rapport with their child using Hetherington and Clinempeel’s (1992) 9-item measure. An example item was “How well do you get along with (child’s name)?” with responses ranging from 1 (not at all) to 5 (a lot). This scale displayed acceptable internal consistency (Cronbach’s alpha = 0.83). For analytic clarity, we presented mean change in the rapport variable across time as a moderator at the between-person level. The average level of rapport for this study was 4.50 (SD = 0.37). Correlations of all study variables are presented in Table 1.

Table 1.

Correlations of Study Variables (N = 316)

Study Variables	1	2	3	4	5	6	7	8	9	10	11
1. Time in study	---	0.02	0.26**	0.20**	0.25**	0.16**	0.19**	0.03*	0.11**	0.03	0.03*
2. Dating initiation		---	0.54**	−0.54**	−0.35**	0.22**	0.17**	0.90**	−0.05**	0.06**	0.02
3. Time in dating relationship			---	−0.29**	−0.19**	0.10**	0.19**	0.56**	0.00	0.09**	0.05**
4. Breakup of dating relationship				---	0.65**	−0.12**	−0.09**	−0.48**	0.04**	−0.01	0.02
5. Time single after breakup					---	−0.08**	−0.06**	−0.32**	0.05**	−0.01	0.01
6. Initiation of subsequent dating relationship						---	0.73**	0.19**	0.00	0.01	−0.01
7. Length of subsequent dating relationship							---	0.17**	0.02	0.02	0.02
8. Relationship quality of dating relationships								---	−0.05**	0.08**	0.02
9. Children’s externalizing behaviors									---	0.28**	−0.06**
10. Children’s internalizing behaviors										---	−0.02
11. Children’s prosocial behaviors											---

Dating initiation, breakup of dating relationship, and initiation of subsequent dating relationship are dichotomous variables (0 = event did not happen; 1 = event happened). Children’s behaviors are measured on a scale of 1–3, with higher scores indicating an increased presence of those behaviors

Analytic Approach

To address the study hypotheses, we employed hierarchical linear modeling techniques (HLM) using HLM software (Raudenbush & Bryk, 2002). HLM accounts for missing data using Full Maximum Likelihood. For the current study, no data were missing at the between-person level for this dataset. Missing data occurred at random only when mothers failed to complete a monthly survey. The independent variables of this study (i.e., dating transitions) are discontinuous in that mothers reported the exact date of a dating transition; unlike a continuous measurement (i.e., Month 1, Month 2, etc.), these exact dates vary across the study, as mothers completed monthly surveys on different days over the course of the study. As such, we used a discontinuous time, longitudinal approach that can measure multiple intercepts and multiple slopes, discussed by Singer and Willett (2003). With this approach, multiple intercept and slope variables are added at the within-person level to capture changes in time with a dependent variable, which also varies at the within-person level. This approach is acceptable when the different intercepts and slopes follow a linear order. For the current study, we have three intercept variables and three slope variables at level 1: event when mothers enter either a casual or serious dating relationship (intercept), time in that relationship (slope), the event of a breakup (intercept), time of being single after breakup (slope), the event of entering a subsequent dating relationship (intercept), and time in the subsequent dating relationship (slope). This approach allows us to measure a change in intercept and slope of children’s behaviors at the event of a dating transition. Given that ~90% of mothers reported changes in their children’s behavior across the study, it is possible that different intercepts and slopes may provide an explanation for this variability. At level 2, we add child-mother rapport (grand mean centered) on each intercept and slope variable to capture the moderating effect of this variable between dating transitions and children’s behaviors (see Eq. (1)). We employed the same model for each dependent variable: children’s internalizing, externalizing, and prosocial behaviors. Additionally, time since divorce filing was also included as a level-1 predictor to control for changes in children’s behaviors independent of mothers’ relationship status.

$$\begin{gathered} C{B}_{ij}=B_{00}+B_{10}\left(TIM{E}_{ij}\right)+B_{20}\left(\mathit{\text{DATINGSTAR}}{T}_{ij}\right) \\ +B_{21}\left(\mathit{\text{MCRAPPORT}}*\mathit{\text{DATINGSTAR}}{T}_{ij}\right)+B_{30}\left(\mathit{\text{DATINGLENGT}}{H}_{ij}\right) \\ +B_{31}\left(\mathit{\text{MCRAPPORT}}*\mathit{\text{DATINGSTAR}}{T}_{ij}\right) \\ +B_{40}\left(\mathit{\text{BREAKU}}{P}_{ij}\right)+B_{41}\left(\mathit{\text{MCRAPPORT}}*\mathit{\text{BREAKU}}{P}_{ij}\right) \\ +B_{50}\left(\mathit{\text{BREAKUPTIM}}{E}_{ij}\right)+B_{51}\left(\mathit{\text{MCRAPPORT}}*\mathit{\text{BREAKUPTIM}}{E}_{ij}\right) \\ +B_{60}\left(\mathit{\text{NEXTDATIN}}{G}_{ij}\right)+B_{61}\left(\mathit{\text{MCRAPPORT}}*\mathit{\text{NEXTDATIN}}{G}_{ij}\right) \\ +B_{70}\left(\mathit{\text{NEXTDATINGTIM}}{E}_{ij}\right)+B_{71}\left(\mathit{\text{MCRAPPORT}}*\mathit{\text{NEXTDATINGTIM}}{E}_{ij}\right) \\ +B_{80}\left(\mathit{\text{RELQUALIT}}{Y}_{ij}\right)+B_{81}\left(\mathit{\text{MCRAPPORT}}*\mathit{\text{RELQUALIT}}{Y}_{ij}\right) \\ +r_0+r_1\left(TIM{E}_{ij}\right)+e_{ij} \end{gathered}$$ (1)

Accordingly, CB_ij refers to the measure of children’s behaviors (internalizing, externalizing, or prosocial behaviors); TIME_ij refers to time in the study for mother i at time j; DATINGSTART_ij refers to the within-person (intercept) effect of starting a dating relationship for mother i at time j; DATINGLENGTH_ij refers to the within-person (slope) effect of staying in a relationship for mother i at time j; BREAKUP_ij refers to the within-person (intercept) effect of breaking up for mother i at time j; BREAKUPTIME_ij refers to the within-person (slope) effect of time since breakup for mother i at time j; NEXTDATING_ij refers to the within-person (intercept) effect of starting a subsequent dating relationship for mother i at time j; NEXTDATINGTIME_ij refers to the within-person (slope) effect of the subsequent dating relationship for mother i at time j; RELQUALITY_ij refers to the relationship quality for mother i at time j; and MCRAPPORT_ij represents that average level of rapport between children and mothers over the course of the study. This equation illustrates six cross-product terms, with MCRAPPORT interacting with each dating transition. For example, MCRAPPORT*DATINGSTART_ij captures the moderating effect of child-mother rapport and the event of mothers starting a dating relationship for mother i at time j on the dependent variable. The residual components are represented by e_ij. The intercept and slope (i.e., TIME) include random effects, whereas all other level-2 equations include fixed effects (Singer & Willett, 2003). We report Akaike and Bayesian information criteria (AIC, BIC) for model fit, and R² to determine variability explained by each model as described by Woltman and the colleagues (2012). Prior to analyses, we examined differences between mothers (with dating experience, n = 270) in which this divorce was their first divorce (n = 243) vs. mothers who had previously divorced (n = 73) using the same models aforementioned. We did not find any significant differences based on divorce status for children’s problem behaviors, and therefore used the full sample of mothers (N = 316) for all analyses. Additionally, we ran all models with the following control variables: mothers’ age, mothers’ income, mothers’ education, length of previous marriage, length of separation, children’s age, children’s gender, and children’s age by gender. In all models, none of the control variables were significant. Therefore, we did not include these variables in our final models. Including these variables could also risk multicollinearity issues.

Results

The first hypothesis predicted that post-divorce dating transitions would be associated with changes in children’s externalizing, internalizing, and prosocial behaviors. Results for this hypothesis are presented in the top of Table 2 (the within-person effects). According to these results, no post-divorce dating transitions (entering a dating relationship, exiting a dating relationship, and/or starting a subsequent dating relationship) were associated with children’s externalizing, internalizing, and prosocial behaviors. Mothers’ dating relationship quality was negatively associated with children’s internalizing behaviors and time in the study was associated with fewer externalizing behaviors.

Table 2.

Examining changes in children’s behaviors during mothers’ post-divorce dating (N = 316)

Independent variable	Externalizing Behaviors	Internalizing Behaviors	Prosocial Behaviors
Slope	−0.03 (0.01)***	0.00 (0.01)	−0.01 (0.01)
Within-person effects
Dating initiation	−0.02 (0.02)	0.01 (0.02)	0.00 (0.01)
Time in dating relationship	0.00 (0.01)	−0.01 (0.01)	0.00 (0.01)
Breakup of dating relationship	−0.03 (0.02)	−0.03 (0.01)	−0.02 (0.01)
Time single after breakup	−0.01 (0.03)	0.01 (0.02)	0.03 (0.02)
Initiation of subsequent dating relationship	0.00 (0.01)	−0.01 (0.01)	0.01 (0.01)
Length of subsequent dating relationship	0.03 (0.02)	0.03 (0.03)	−0.02 (0.02)
Relationship quality	0.00 (0.00)	0.00 (0.00)*	0.00 (0.00)
Between-person effects
Dating initiation*Mother-Child Rapport	−0.03 (0.05)	−0.01 (0.04)	0.04 (0.04)
Time in dating relationship*Mother-Child Rapport	−0.04 (0.02)	0.00 (0.02)	0.05 (0.02)**
Breakup of dating relationship*Mother-Child Rapport	−0.11 (0.04)**	−0.06 (0.04)	0.05 (0.04)
Time single after breakup*Mother-Child Rapport	0.03 (0.10)	0.06 (0.07)	−0.05 (0.06)
Initiation of subsequent dating relationship*Mother-Child Rapport	−0.02 (0.04)	0.02 (0.03)	−0.04 (0.03)
Length of subsequent dating relationship*Mother-Child Rapport	0.10 (0.07)	−0.10 (0.06)	−0.03 (0.04)
Relationship quality*Mother-Child Rapport	−0.01 (0.01)	0.00 (0.00)	0.00 (0.00)
Goodness-of-fit
AIC	−1064.13	−2737.88	−3798.41
BIC	−1054.14	−2727.89	−3788.41
R 2	0.06	0.07	0.04

Statistics are standardized beta coefficients and presented as B(SD)

^***p < 0.001,

^**p < 0.01,

^*p < 0.05

The second hypothesis predicted that child-mother rapport would moderate the relationship between dating transitions and children’s behaviors. Results for these analyses are presented in the bottom of Table 2 (the between-person effects). These results reveal that child-mother rapport moderated the relationship between children’s externalizing behaviors and the event of breakup (i.e., intercept) of a dating relationship after divorce. This interaction effect is presented in Fig. 1. According to this figure, children experienced fewer externalizing behaviors at the time of a post-divorce dating breakup when mothers reported high rapport with children compared to when rapport is low. The amount of variance explained by this model was 6%. Besides this interaction, child-mother rapport did not moderate any of the other dating transitions for children’s externalizing behaviors.

Fig. 1.

This graph depicts the interaction of child-mother rapport and dating breakup for children’s externalizing behaviors. Children reported significantly less externalizing behaviors when they had high levels of rapport with their mothers at the time that mothers experilenced the breakup of a post-divorce dating relationships compared to when they had lower levels of rapport with their mothers

Additionally, no significant interacting effects were found for child-mother rapport between post-divorce dating transitions and internalizing behaviors, but a significant interaction was identified between child-mother rapport and time in a dating relationship for children’s prosocial behaviors (Fig. 2). Thus, the strength of the effect of the duration of a dating relationship on prosocial behavior is dependent upon the child-mother rapport. This means, assuming mothers’ dating time is the same (or controlled for), children with a high rapport with their mothers exhibited more prosocial behaviors than peers who has a lower rapport. No other interactions were identified between dating transitions and prosocial behaviors. The amount of variance explained for the model predicting prosocial behaviors was 4% and the amount of variance explained by the model predicting internalizing behaviors was 7%.

Fig. 2.

This graph portrays the interaction of child-mother rapport with time when mothers are dating and children’s prosocial behaviors. When mothers remained in dating relationships, children exhibited more prosocial behavior as the time went by regardless of child-mother rapport. When time is controlled for, children displayed more prosocial behavior when their rapport with mother was high than when the rapport was low

Discussion

This study aimed to understand how post-divorce dating transitions impact children’s behaviors. Opposite to our first hypothesis, the results show that none of the post-divorce dating transitions significantly predicted children’s externalizing, internalizing, or prosocial behaviors. However, two significant interaction effects were identified, lending some support for our second hypothesis. Child-mother rapport moderated the link between post-divorce breakup and children’s externalizing behaviors, and the link between time in a dating relationship and children’s prosocial behaviors. These results demonstrate that various post-divorce dating transitions as a whole appear to have little direct effect on children’s adjustment, but rapport between children and mothers can moderate the effect of a few post-divorce transitions on children’s behaviors; yet, such moderation does not apply consistently across all transitions.

We opt to focus on the second hypothesis first as the results yielded two significant interaction effects. We found the effect of ending a post-divorce dating relationship on children’s externalizing behaviors varies by the child-mother rapport. Children having higher rapport with their mother exhibited fewer externalizing behaviors than children having lower rapport with their mother when their mother’s dating relationship dissolved. Following divorce, breakup with a dating partner could compound mother’s adjustment, putting her well-being at stake. This potential decline in a mother’s psychological health may hamper a child’s adjustment. However, high family harmony, characterized by more agreement, greater cohesion, and less conflict, is predictive of better maternal health conditions (Grein & Glidden, 2015). It is possible that, at the time of relational breakup, a strong child-mother relationship helps keep mothers from being vulnerable. Then the resultant mothers’ preserved health state and the supportive child-mother relationship together serve as protective factors, reducing the impact of breakup on children and keep at bay their externalizing behaviors such as aggression and defiance.

Also, child-mother rapport interacted with the time in a dating relationship on children’s prosocial behaviors. It represents that, during the same length of time in mothers’ dating, children having better rapport with their mother exhibited more prosocial behaviors than those with lower child-mother rapport. A healthy parent-child relationship where closeness and autonomy co-exist has been shown to be predictive of a children’s empathy and prosocial behaviors (Yoo et al., 2013). Ferreira et al., (2016) demonstrated that higher quality father-child dyads and teacher-child dyads were associated with more prosocial behavior in children. Feeling emotionally secure toward marital discord is found to be positively associated with prosocial behavior for a child (McCoy et al., 2009). So, coupled with the literature, the moderation effect denotes that a high child-mother rapport could engender and preserve a sense of emotional security for children, and the resulted comfort and assurance strengthen the positive effect of mothers’ dating time on children’s prosocial behavior. Although the moderating role of child-mother rapport does not show up pervasively as anticipated in current research, the two yielded interaction effects still corroborate that a high child-mother relationship can work as a buffer for children at times of family changes such as post-divorce dating transitions. The non-significant interaction effects of child-mother rapport on other transitions implicates that child-mother rapport may be more likely to operate as a mediator than as a moderator in the context of family adjustment following divorce. We suggest that future research examine both direct and indirect effects of the child-parent relationship on child’s adjustment to stressful events following divorce.

Regarding the first hypothesis, we did not find support for the direct effects of post-divorce dating transitions for children’s behaviors. A few potential explanations could facilitate sensemaking of the non-significant findings. First, the lack of significant effects of all studied post-divorce dating transitions might be attributed to the characteristics associated with these relational experiences. As presented earlier, there are discrepancies between dating transitions and repartnering. When the relationship moves to repartnering, children likely have more time with their mother and her romantic partner as triadic activities among children, mothers, and her partners increase. This condition makes it easier for children to be familiar with and form a proximal relationship with their mother’s partner. In contrast, whether and to what extent children are informed of and involved in dating transitions are at mothers’ discretion. Nixon and Hadfield (2018) argued that mothers were cautious about when to introduce their dating partner to children. If the introduction occurs prematurely or is not handled well, it would likely evoke unpredictable instability and stress for the family (Beck et al., 2010; Hetherington & Kelly, 2003). Gray et al. (2016) contended that single parents have to balance a wide range of considerations (e.g., the well-being of child, the level of commitment with the dating partner) when deciding the timing and extent of involvement of children in their dating life. This wariness could prompt single mothers to keep their dating transitions, particularly at the early stage of dating, outside their children’s view and knowledge. Consequently, children may even be unaware of whether or not their mother is in a romantic relationship; or they know of nothing beyond the fact that the mother is dating or breaks up with someone. In this sense, until a committed relationship is well established and mothers feel confident in the compatibility between the partner to both mothers and child, the chances are that children have nothing more than a limited knowledge of their mother’s dating transitions. Pertinent questions such as when mothers started dating, whom they were dating, how long the dating lasted, and when the breakup happened are probably beyond children’s scope of abilities to answer due to the inadequacy of information and involvement opportunity afforded by some mothers. Accordingly, mothers’ dating transitions are unlikely to be regarded or experienced as emotion-laden events by children at this point, a situation that presumably has little power to directly impact children’s emotional security. When the sense of emotional security remains relatively intact, there will be few tangible adjustments in children’s behaviors (Davies & Cummings, 1994). In the current research, this is reflected by the non-significant effects of dating transitions on the sampled children’s externalizing, internalizing, and prosocial behavior.

Comparable to children’s awareness of mothers’ dating experiences, children’s developmental characteristics could also play a critical role. The age range of the sampled children is 4–12 (SD = 1.97). Mazur (1993) identified salient developmental differences in children’s attitudes, perceptions, and reasoning regarding marriage, divorce, and remarriage. Specifically, in portraying what characterizes a good ex-spouse and stepparent, kindergartners relied on observable actions and appearances; 2nd to 4th graders resorted to concrete skills and abilities whereas a few children aged 9–10 were able to consider psychological attributes and feelings. The disparities in cognitive and social functions such as perspective taking, abstracting thinking, and verbal skills (Su & Teoh, 2018) among children of varying developmental stages contribute to nuanced opinions they hold toward mothers’ relational experiences as well as variations in abilities to evaluate the quality of mothers’ dating relationship and compatibility of the partner for the current family life and structure. In cases where children were aware of mothers’ dating experiences and had adequate opportunities to associate with the mother’s partner, younger groups (e.g., aged 4–6) might be unable to understand the meaning of these transitions due to the lack of related cognitive capacity. For those who are old enough to make sense of mothers’ romantic needs and motives, they might opt to normalize mothers’ dating transitions, viewing forming or discontinuing a relationship as part of mother’s ordinary life. Per emotional security theory (Davies & Cummings, 1994), children’s appraisal of family events and their capabilities of handling the associated stress determine their emotional security and the ensuing adjustment. When children do not conceptualize mothers’ dating transitions as threatening or detrimental, their adjustment are unlikely to be disrupted. We recommend that future studies, when dealing with a large sample of children with a wide age range, consider dividing the sample into small groups and examine the potential group differences as a function of developmental factors. This effort could help maximize the value of data and yield findings uncovering effects of children’s developmental characteristics in relation to their adjustment following family dissolution.

Alternatively, the absence of sizeable effects of post-divorce dating transitions may reflect possible indirect pathways through which transitions impact children adjustment through potential mediators. For example, inter-parental conflict, parenting, and parent’s health conditions were found to mediate the link between divorce and children’s adjustment (Lansford, 2009). A mother’s dating relationship can influence the biological father’s emotions that leads to changes in paternal involvement and quality of co-parenting, which can either strengthen or weaken children’s emotional security depending on the paternal father’s response to mothers’ dating (Davies & Cummings, 1994; Hadfield & Nixon, 2017). Maternal well-being may play a mediating role as well. For example, mothers’ increased depressive symptoms are associated with more externalizing behaviors of children of divorced parents (Weaver & Schofield, 2015). If the experiences of dating transitions (e.g., breakup) are unpleasant, it could put mothers’ psychological health at risk. The psychological vulnerability could be mirrored by children directly or it could impede mothers’ parenting and then contribute to children’s maladjustment. Meanwhile, staying in a promising dating relationship might evoke greater involvement of the dating partner within the family. The forms of involvement could range from the filling of a parenting role to the provision of support for mother’s and/or children’s adjustment, both of which are able to exert more direct effects on children’s post-divorce behaviors. Given the presence of these in-between mediators, it is possible that mothers’ dating transitions trigger changes in certain determinants of children’s adjustment instead of influencing it directly. Obviously, each of these proposed explanations entails verification from future research.

We also found two significant effects that do not directly pertain to our major research goals but could bear value for future studies. First, time is associated with declines in child’s externalizing behaviors. It is not uncommon for children’s problem behaviors to decrease over time as children adjust to new family dynamics after divorce (Hetherington & Kelly, 2003; Wang & Amato, 2000). A failure model (Capaldi, 1992) proposed that externalizing behavior predates internalizing behavior, operating as a causal trigger, which is supported by research (Burt & Roisman, 2010; Gooren et al., 2011). Maybe the improvement of externalizing problems along the timeline contributes to the reduction of internalizing problems, which partially counteracts the ameliorative effect of time on emotional behaviors. Also, the quality of mothers’ dating relationships predicted declines in children’s internalizing behaviors. This finding suggests that mothers’ supportive dating relationships generate more emotional than behavioral benefits for children. For example, high quality dating relationships sustained and boosted mothers’ mental health (Langlais et al., 2016a), which is associated with lower risk of children feeling depressed and anxious. In contrast, lower quality dating relationships, which is characterized by inadequate support and recurring and intense conflicts, could be a source of stress for children and take a toll on children’s emotional security and contribute to the increase of their internalizing behaviors.

It is worth noting that all significant effects identified in the study are applied to either externalizing or internalizing behaviors. While certain research (Basten et al., 2016; Hinnant & El-Sheikh, 2013) identified the co-occurrence of externalizing and internalizing behaviors in young children, especially in at-risk group who having a higher rate (Willner et al., 2016), these two types of behaviors are unarguably distinct. That being said, the research results still pose a challenge of interpretation: how do we explain that the significant effects take place on either externalizing or internalizing behavior instead of both of them? Few studies have examined and uncovered the mechanisms of how divorce and divorced-related events induce and impact externalizing and internalizing behaviors of children (Oh et al., 2020). The share-risk hypothesis (Angold et al., 1999) proposed that the co-occurrence of both externalizing and internalizing behaviors is a function of common risk factors. In our case, it is mothers’ lower dating relationship quality and lower child-mother rapport following divorce. However, as Oh et al., (2020) pointed out, the validity of shared-risk hypothesis has not been fully established. Upon checking findings regarding externalizing and internalizing behaviors across studies, they underscored the remaining unsettled entanglement of externalizing and internalizing behaviors. Does one type of behavior lead to the occurrence of another type? In what situations do they emerge con-currently? Is there a common source that results in both types of behaviors? Future research is recommended to keep these unaddressed questions as part of the foci of their scientific investigation.

Strengths, Limitations, and Conclusions

This study is one of few studies that examines the impact of mothers’ post-divorce dating transitions for children’s adjustment. Previous studies in this territory have primarily focused on cohabitation and remarriage after divorce. By studying post-divorce dating transitions, the present research contributes to the literature by illustrating that post-divorce dating transitions are not comparably powerful as cohabitation and remarriage in terms of impacts on child’s emotional security and subsequent adjustment. Another strength of this study stems from the methodology. As a longitudinal study, the current study measures not just one but multiple dating statuses – initiating dating, maintaining of dating, transition from dating to breakup, maintaining of breakup, and resuming another dating relationship – all of which were measured monthly for a period of 24 months. Because these dating transitions could occur and stop for an unpredictable duration of time at any time during the study, by using HLM we managed to capture the discontinuous nature of the data. With the accurate detection of the intercept and slope for each dating transition, this study has yielded a dataset that can help paint a relatively comprehensive picture with regard to how correlated post-divorce dating transitions individually and collectively are with children’s adjustment.

While possessing the promise of contributing to the literature, the current research is not without its limitations. First, there were limitations concerning our sample. Nine hundred and eighty-seven mothers were originally contacted to participate in a get acquainted visit, and although more than 80% of mothers who agreed to this meeting participated, the mothers who did not participate in the get acquainted visit may be the families needing the most help and support after divorce. Second, the measurement of all study variables come from mothers’ reports, a situation that is susceptible to the social desirability effect (the mean score of child-mother rapport was 4.5 out 5). Additionally, 2.2% of mothers reported no externalizing or internalizing behaviors for their children over the course of the study and an additional 2.5% reported no variability in their children’s behaviors over the course of the study, which could be a reflection of maternal bias in favor of the target child. Considering these factors could bias the results, the inference on the generalizability of the research findings to a larger population should be drawn with caution. Future studies should incorporate voices of multiple informants to produce a more complete assessment of children’s emotions and behaviors.

Although we examined demographic variables of children, a limitation of this study is that we did not control for other variables that might be associated with children’s divorce experiences. Probably the most consequential variable is the children’s awareness of mother’s dating transitions. If children are not informed of whether mothers are in a dating relationship or not, or mothers’ efforts to share (or not share the relationship) with their children are not captured, it is unlikely that significant effects on children’s behavior will be found. When children are aware of mothers’ dating transitions, there are two possible situations. Mothers impart their relational status to children and keep them updated. Or mothers choose not to tell children but for whatever reason children find out the mother is in some types of dating transition. The different scenarios, when interacting with children’s attributes, particularly their ways of conceptualizations of mothers’ decisions to move in and out of romantic relationships as partially a function of their developmental age (discussed above), could impact children’s adjustment. Similarly, Lansford (2009) suggested children’s gender and race/ethnicity is associated with children’s post-divorce adjustment. As mentioned previously, the inclusion of these variables revealed no significant effects, and more importantly, the main findings of each model remained non-significant. Next, this study focused on the first two years after families filed for divorce. Some mothers may have started dating during the separation period, which could alter how dating transitions influence children, as dating relationships that formed before the divorce was finalized may impact children differently compared to dating relationships after a divorce was finalized. However, we could not control for relationships that started before baseline based on study design. Additionally, in terms of transitions, we were unable to account for the instigator of various relationship transitions, such as who sought out a dating relationship (the mother or the dating partner) as well as who sought the end of a dating relationship. It is possible that the impact of dating relationships may depend on who initiated the dating transition. Although the current study uses longitudinal data, more research is needed to understand the bidirectional associations between post-divorce dating transitions and children’s externalizing, internalizing, and prosocial behaviors. It may be possible that children’s behaviors may prompt mothers’ post-divorce dating transitions. There is evidence that some children make parental dating after divorce more difficult for parents (Hetherington & Kelly, 2003), but more research is needed to test whether children’s problem behaviors would also prompt other relationship transitions.

It is acknowledged that repartnering embodies a stronger commitment than post-divorce dating. However, for most divorced persons, dating after divorce is an unavoidable step before proceeding to repartnering. Many single parents, while possessing the desire to resume seeking that “right” person, hesitate to do so partially due to the tension between parenting and dating. The current research affords some alleviating evidence to those parents that they could brave dating after divorce without worrying about the impact on their children, particularly if they have strong relationship with their children.

Highlights.

• Prior to cohabitation and remarriage, divorced parents are likely to go through multiple post-divorce dating transitions, which are assumed to be able to shape family adjustment.

• Results reveal that child-mother rapport moderated the relationship between children’s externalizing behaviors and the breakup of a dating relationship after divorce, and the relationship between children’s prosocial behavior and the time of a dating relationship.

• Other post-divorce dating transitions did not predict changes in children’s behaviors.