Being separated from their parents can be a dramatic change in the lives of young children, requiring them to adjust psychosocially to the physical absence of their parents. This adjustment can go well but can also sometimes lead to behavioral difficulties in the form of externalizing problems (e.g., hyperactivity symptoms or aggressive behaviors) and/or internalizing problems (e.g., symptoms of depression, anxiety) or reduced socio-emotional skills (Gilliom and Shaw, 2004; Carta et al., 2009).

With regard to child adjustment, Prevatt (2003) conceptualized and evaluated a risk and resilience model based on previous seminal research (Baumrind, 1967; Darling and Steinberg, 1993), in which different parenting behaviors of caregivers (parents and/or grandparents) protect or jeopardize child adaptation. In the meta-analysis by Pinquart (2017a, 2017b), positive parenting, including parenting dimensions such as warmth, behavioral control, autonomy granting, and the authoritative parenting style (i.e., affectively responsive parenting with high degrees of democratic participation and child regulation, Wu et al., 2002) are negatively associated, albeit in a small-sized manner, with child difficulties, while the associations for negative parenting like the parenting dimensions harsh control, psychological control, and the authoritarian parenting style (i.e., hostile parenting with high levels of unjustified punitive control, Wu et al., 2002), are the reverse.

Grandparents are ubiquitously the primary caregivers in families where both parents migrate for the sake of their jobs. They are especially influential in the immediate and direct environment around the children, and grandparenting thus plays a crucial role in young left-behind children’s adjustment. Many existing research on left-behind children and intergenerational caregiving in China has largely emphasized risks (such as prolonged separation, inconsistent caregiving, and overindulgent parenting) and has often portrayed grandparent care primarily as a source of problems (Gu, 2022; Shan et al., 2025). Indeed, previous studies have shown that children who are raised entirely by grandparents because of their parents’ migration generally display more maladaptive and fewer prosocial behaviors compared to parent-raised non-left-behind children (Fan et al., 2010; Fellmeth et al., 2018; Hu et al., 2020), this does not mean grandparent childrearing per se undermines child development. Some researchers argued that these differences are possibly related to poverty and parental absence instead of whether the caregiver is a grandparent (Li et al., 2019; Yang and Liu, 2020). A limited number of studies have examined how grandparenting dimensions or styles are related to child adjustment outcomes (Weissman et al., 2016; Hou et al., 2019; Li et al., 2019; Yang and Liu, 2020). As summarized above, prior work has suggested that both supportive and suboptimal grandparenting practices may be linked to children’s emotional and behavioral problems, but the findings regarding grandparental warmth are mixed. For instance, Li et al. (2019) found that suboptimal parenting practices by both grandparents and parents increased the risk of emotional and behavioral problems in Chinese adolescents, and reported a positive association between grandparental warmth and child difficulties, whereas other studies have documented negative associations between grandparental warmth and child difficulties (e.g., Hou et al., 2019; Yang and Liu, 2020). Thus, existing evidence on the role of grandparental warmth in child adjustment is inconsistent. While this risk-focused work is important, the risk–resilience literature also highlights the need to identify protective processes that help children adapt under adversity. In the present study, we therefore adopt a strengths-based perspective and focus on positive/authoritative grandparenting (alongside distant maternal responsivity) as a potential protective caregiving resource in high-risk left-behind families, rather than attempting to model the full range of grandparenting styles. Based on the literature, we formulated the following hypothesis:

In face-to-face contexts, when only parenting by mothers is examined (and parenting by fathers is not), the associations generally remain the same (negative associations between positive parenting and child difficulties, and positive associations between negative parenting and child difficulties) as the meta-analysis findings by Pinquart (2017a, 2017b). Moreover, more positive and less negative mothering has been found to be associated with stronger prosocial and emotional competence of the child (Garner, 2006; Wang et al., 2006; Pastorelli et al., 2016).

As with general parenting, it is assumed that the special context of distant mothering also has an impact on child functioning (Madianou and Miller, 2011; Nichols and Selim, 2022; Tang et al., 2022). To be able to operationalize distant mothering, new measures for distant parenting have been developed (Weisskirch, 2009; Jensen et al., 2021; Liang and Van Leeuwen, 2024). Within our larger research project, an initial qualitative inquiry with migrant parents (Liang et al., 2023) identified six recurrent dimensions of mobile phone parenting. A subsequent quantitative study (Liang and Van Leeuwen, 2024) validated this six-factor structure (i.e., responsivity, autonomy granting, psychological control, proactive control, directive control, and harsh punitive control) in a sample of LBC’s parents. In that study, responsivity, autonomy granting, and proactive and directive control were positively intercorrelated, whereas psychological control and harsh punitive control clustered together and were negatively related to the more positive dimensions. This pattern mirrors distinctions in the general parenting literature between more supportive, structured practices and more intrusive or harsh forms of control.

From this perspective, distant responsivity, autonomy granting, and proactive and directive control can be viewed as reflecting relatively positive and structured distant mothering, whereas psychological and harsh punitive control reflect more negative distant mothering. However, despite the initial conceptualization and psychometric validation of these six dimensions, little is known about how they are empirically related to child prosocial behavior and difficulties, or whether distant mothering makes a unique contribution to left-behind children’s socio-emotional outcomes beyond grandparenting. Therefore, in the present study distant mothering is our primary focal construct, whereas grandparenting is included as a broader indicator of positive caregiving in the home context rather than being decomposed into multiple dimensions. Based on the literature on general mothering and child adjustment and on our prior work on distant parenting, we formulated the following hypothesis:

Coparenting relationship quality refers to the extent to which parenting partners coordinate their childrearing efforts and support or undermine one another in parenthood (McHale, 1997; Feinberg, 2003). Boričević Maršanić and Kušmić (2013) concluded in their review that coparenting is a multidimensional concept including discrepancies in parental involvement and supportive/hostile-competitive aspects. In the comprehensive model proposed by Feinberg et al. (2012), coparenting relationship quality is defined in terms of coparenting agreement, coparenting closeness, exposure of child to conflict, coparenting support, coparenting undermining, endorsement of partner’s parenting, and division of labor. While most studies examine parental coparenting relationships within Chinese nuclear families (McHale et al., 2000; Ren and Xu, 2019; Fan and Chen, 2020; Liu et al., 2021; Wang et al., 2022) there is an increase in research into the mother-grandparent coparenting relationship (Li and Liu, 2019; Li et al., 2020; Luo, 2022; Bai et al., 2023; Xu et al., 2023).

The specific associations between coparenting relationship quality and child adjustment have been demonstrated in two meta-analytic studies, in which fewer children’s problem behaviors and better social functioning were significantly associated with a favorable coparenting relationship, but the effect sizes were small (Teubert and Pinquart, 2010; Zhao et al., 2022). The associations have been especially confirmed in Chinese families where fathers and mothers jointly parent their preschool-age children (Lam et al., 2018; Ren and Xu, 2019). Also, more recent studies on children in China have explored the associations between intergenerational coparenting relationship quality and children’s socio-emotional outcomes. For example, the study by Li and Liu (2019) on Chinese young children suggested that mother-grandparent coparenting relationship quality positively predicted the social competence of preschoolers. Another study further found that the effects of intergenerational coparenting relationship quality on social competence of preschoolers differed in strength between mother-maternal grandmother and mother-paternal grandmother coparenting patterns, with maternal grandmothers’ perception of coparenting relationship quality as a stronger predictor (He et al., 2023). While these studies have provided emerging evidence in urban co-residing family members, it is still unknown whether the dynamics of distant coparenting are similarly linked to the social–emotional outcomes of rural left-behind children. This leads us to the following hypothesis in the current study:

In the ecological model of coparenting, Feinberg (2003) posits a high-quality coparenting relationship as a moderator to buffer or protect children from the influence of adverse contexts (e.g., negative parenting). It is also reasonable to assume that a good coparenting relationship can facilitate the benefits children derive from nurturing environments (e.g., positive parenting). Indeed, initial findings have shown the moderating role of the coparenting relationship quality in the association between (grand)parenting and child outcomes (Scrimgeour et al., 2013; Latham et al., 2017; Yan et al., 2021; Zhang et al., 2022). For instance, as regards the facilitative model, Scrimgeour et al. (2013) found that in 58 two-parent American families, the interaction of maternal inductive reasoning (assessed when the child was 24 months) and cooperative coparenting (assessed when the child was 42 months) positively predicted the prosocial behavior of four-year-old children. To our knowledge, the moderating role of mother-grandparent coparenting relationship in the link between (grand)parenting and early childhood adjustment has rarely been investigated. Nevertheless, we can draw inspiration for a hypothesis in the current study, from a cross-sectional study on parent-grandparent coparenting in a sample of Chinese adolescents. Zhang et al. (2022), who included both coparenting relationship and parent-grandparent relationship as moderators, reported that the positive association between low-supportive grandparenting and adolescent depressive symptoms was significant when the adolescents perceived high quality of the parent-grandparent relationship, but at the same time less coparental cooperation. We thus hypothesize:

The present cross-sectional study examined how three interrelated family subsystems (positive grandparenting, mobile-phone-based distant mothering, and the quality of the mother-grandparent coparenting relationship) are associated with child adjustment (i.e., prosocial behavior and difficulties) among Chinese families with young left-behind preschoolers in rural areas. Extending prior research on Chinese intergenerational parenting, which has mainly focused on grandparent caregiving in co-residing three-generation households (See Bai et al., 2023, for a review), we focus on families in which both parents have migrated, grandparents are the primary daily caregivers, and mothers engage in parenting from a distance via mobile phones. We conceptualize maternal parenting at a distance as a full parenting context, assessing multiple dimensions of distant mothering (responsivity, autonomy granting, proactive and directive control, psychological control, and harsh punitive control) alongside positive grandparenting.

This study may contribute in two main ways. First, it advances theoretical work by synthesizing and extending existing models. We integrate the risk/resiliency model of child adjustment (Prevatt, 2003) and the ecological model of coparenting (Feinberg, 2003) within a family systems framework (Minuchin, 1985; Cox and Paley, 1997), and extend these models to a translocal context in which caregivers share childrearing responsibilities despite long-term physical separation and intensive media use. In particular, we investigate whether the quality of the mother-grandparent coparenting relationship moderates associations between (grand)parenting behaviors and child outcomes in these digitally connected but geographically separated families. In doing so, we address a key gap in previous LBC-related studies, which have largely focused on grandparenting alone and have overlooked the contribution of distant parenting and its interaction with grandparent caregiving (Hou et al., 2019; Yang and Liu, 2020; Pan et al., 2021).

Second, this study enriches the empirical literature on intergenerational parenting in China by focusing on young LBC in an economically less developed rural region and by employing a multi-informant triadic design. In contrast to most previous intergenerational coparenting studies that relied on urban families and single- or two-informant designs (Hou et al., 2019; Li et al., 2019; Danielsbacka et al., 2022; Xu et al., 2022; Zhang et al., 2022), we combine reports from preschool teachers (child adjustment), co-resident grandparents (grandparenting and coparenting), and migrant mothers (distant mothering and coparenting). The findings can inform family-focused interventions by highlighting the need to consider positive grandparenting, distant mothering practices, and intergenerational coparenting quality together when promoting the well-being of LBC.

Triadic data were obtained from a cross-sectional online/telephone survey administered to teachers from 32 rural preschools, migrant mothers, and grandparents of 185 preschool-age left-behind children in Chuxiong Yi Autonomous Prefecture, Yunnan, a province with lower levels of economic development located in the southwestern region of China. According to official statistics (Chuxiong Prefecture Bureau of Statistics, 2023), approximately 37.0% of the local population belong to ethnic minority groups, predominantly Yi (29.7%); however, we did not record individual participants’ ethnic identity, and the ethnic composition of our sample may therefore differ from the population distribution. The present quantitative study is a component of a larger mixed-methods project, for which an initial exploratory qualitative study on mobile phone parenting and intergenerational coparenting in LBC’s families (Liang et al., 2023) provided in-depth insights that methodologically and theoretically informed the focus of the current quantitative investigation. Inclusion criteria were: families (a) where both parents were physically absent; (b) that were experiencing a three-month-above parent–child separation with at least one child aged three to six, left in the hometown; (c) where the grandparents served as primary caregivers; (d) where the mothers had regular communication with their children through mobile phones; and (e) where both grandparents and mothers were willing and able to participate in questionnaires or interviews. Most of the participating grandparents who primarily co-parented the children with migrant mothers, were female. The sample included 55.14% paternal grandmothers, 25.41% maternal grandmothers, 9.19% paternal grandfathers, and 10.27% maternal grandfathers. Migrant mothers were on average in their thirties (M_age = 31.94, SD = 3.76) and 91.35% of them engaged in occupations such as non-technical, semi-technical, or technical work. About 33.51% of the children were being separated from their mothers for 2 years or more. Table 1 presents extra demographic characteristics of the families involved in this study.

The data collection took place from May 2022 to January 2023. In light of COVID-19 restrictions, all questionnaires were completed remotely to ensure the safety of participants and to adhere to local public health guidelines. During this period, the involved preschools and participants were not severely affected by the coronavirus pandemic and still operated as usual except for stricter control on outsider access and heightened precautions. Mothers and classroom teachers received the links to the electronic questionnaires by messages. Teachers reported the target children’s behaviors on the Strengths and Difficulties Questionnaire online. Mothers completed a demographic questionnaire, the Mobile Phone Parenting Practice Questionnaire, and the Coparenting Relationship Scale. Investigators conducted telephone interviews to collect the answers from a grandparent who was the primary caregiver in each target family on the Strengths and Difficulties Questionnaire, Parenting Styles and Dimensions Questionnaire, and Coparenting Relationship Scale. All participants were compensated for their involvement by remuneration. In the beginning, we received a total of 216 teacher questionnaires. Out of the 18 teacher questionnaires that involved two children from the same families, only one of the children was selected. Among the participants, 22 triads had missingness of either the mother questionnaires or grandparent questionnaires. Consequently, we kept the 185 triads with complete information.

The study was approved and carried out following the ethical requirements of the Social and Societal Ethics Committee of KU Leuven (Approval number: G-2021-4012-R2(MAR)). Informed consent was obtained from all the mothers, grandparents, and teachers.

We tested hypotheses H1, H2, and H3 with hierarchical multiple regression analyses. We sequentially entered positive grandparenting, six distant mothering dimensions, and intergenerational coparenting relationship quality into the models with the covariates to examine to what extent these added variables explained variance in child adjustment. The same models were tested twice, once for child difficulties and once for prosocial behavior as dependent variables.

Using the “processR” package in R (Hayes, 2017), we also examined two-way interaction effects (i.e., positive quality of grandparenting × coparenting, and maternal parenting dimensions × quality of coparenting relationship) to test hypothesis H4. If a significant moderator was detected, a subsequent post hoc probing method was carried out to examine and interpret the interaction (Aiken et al., 1991). The first step was conducting separate regression analyses for three levels of the moderator (i.e., 1 SD above the mean, mean, 1 SD below the mean) and plotting the three simple regression lines. Next, conditional effects of the independent variables at the high, mean, and low values of the moderators were examined with t tests.

Table 1 provides the descriptive statistics of parents’ and grandparents’ characteristics, covariates, and child outcomes. Spearman correlation was used to examine the associations between child sex (0 = girl, 1 = boy) and other variables while the other bivariate associations among the key variables were tested with Pearson correlations. In Table 2, the correlations among all variables showed that child prosocial behavior was positively correlated to family income, frequencies of father-child telecommunication, positive grandparenting, the quality of mother-grandparent coparenting relationship, maternal responsivity, and maternal autonomy granting with small to medium effect sizes based on Cohen’s criteria (Cohen, 1988). In contrast, child difficulties had a negative, weak-to-medium association with mother’s completed educational years, family income, frequencies of father-child telecommunication, positive grandparenting, coparenting relationship quality, maternal responsivity, maternal autonomy granting, maternal proactive and directive control, and had a positive, weak-to-moderate association with maternal psychological control and harsh punitive control, respectively.

Table 3 presents the results of hierarchical multiple linear regression analyses with positive grandparenting, dimensions of distant mothering, and the quality of mother-grandparent coparenting relationship as the independent variables and child prosocial behavior and difficulties as the dependent variables.

The interactions between coparenting relationship quality and each dimension of distant mothering/positive grandparenting were examined separately for child prosocial behavior and child difficulties (Table 4). Almost all interaction terms between coparenting relationship quality and distant mothering/grandparenting for child difficulties were insignificant. There was only one significant interaction effect, i.e., maternal proactive control × coparenting relationship quality, predicting prosocial behavior (ΔR² = 0.03, β = −0.17, p = 0.013).

A post hoc test was conducted to interpret the significant interaction effect. Figure 1 depicts the simple regression lines regarding associations between maternal proactive control and child prosocial behavior at low (one SD below the mean), medium (mean score), and high (one SD above the mean) quality of mother-grandparent coparenting relationship. Only the slope for families reporting low quality of coparenting relationship was significant (t = 2.38, p = 0.018). The slopes for the mean (t = 1.08, p = 0.282) and low conditions (t = −0.840, p = 0.402) were not significant. A low-quality coparenting relationship can be considered a risk factor for child prosocial behavior in families with low levels of proactive control by migrant mothers, given that in this situation, the level of child prosocial behavior was lowest. When mothers used high levels of proactive control, the levels of coparenting quality seemed to make little difference to child prosocial behavior as there was no difference in levels of prosocial behavior between the three groups at low, medium, and high quality of coparenting.

We found congruent evidence to support the importance of positive grandparenting in child development (Weissman et al., 2016; Li et al., 2019; Yang and Liu, 2020), especially for LBC in China (Hou et al., 2019), with results demonstrating that children showed more prosocial behavior and fewer adaptive difficulties when their grandparents used more positive parenting practices (i.e., a combination of warmth, democratic participation, and reasoning practices). The result aligns with our Hypothesis 1. The role of positive grandparenting for child adjustment in parent–child-separated families remained significant even when other variables regarding distant mothering and coparenting were taken into account. Grandparents are usually the first primary caregivers in families without parental physical presence. This finding aligns with the idea of the importance of the proximal environment and frequent interaction processes for child functioning and development, outlined in Bronfenbrenner’s bio-psycho-social model (Bronfenbrenner and Morris, 1998). In the Chinese cultural context, norms of filial piety and intergenerational obligation (Yeh and Bedford, 2003) also legitimize grandparents’ central caregiving role and authority in the family, which may further strengthen the impact of their positive grandparenting practices on grandchildren’s adjustment. In terms of grandparents’ protective role, their warmth and support can contribute to the psychological resilience of grandchildren and buffer against the influence of adversities in grandchildren’s life, consequently following a decreased risk of socioemotional disorders (Dunifon, 2013; Li et al., 2021). Additionally, from the perspective of the facilitating role of grandparents (Paulus and Moore, 2012), grandchildren might learn to use prosocial skills in family and preschool social occasions from grandparents’ democratic interactions with them and appropriate control by reasoning.

For bivariate associations between distant mothering dimensions and child adjustment without including covariates, as expected, child prosocial behavior was positively correlated with maternal responsivity and maternal autonomy granting. Child difficulties were negatively correlated with maternal responsivity, maternal autonomy granting, and maternal proactive and directive control while positively correlated with maternal psychological control and harsh punitive control. When also considering positive grandparenting and demographic variables, no evidence of hypothesized links between dimensions of distant mothering and child adjustment was detected, except for the finding that maternal responsivity was significantly positively related to child prosocial behavior. The findings provide partial support for our Hypothesis 2.

Although several distant mothering dimensions showed associations in the expected directions at the bivariate level, the zero-order correlations with child outcomes were modest (|r| = 0.06–0.24), which is comparable to effect sizes typically reported for general parenting (|r| = 0.01–0.24; Pinquart, 2017a, 2017b; Wong et al., 2021). Given these small-to-modest effect sizes, the relatively small triadic sample and the number of predictors included in the regression models, the present study may have been underpowered to detect unique effects of autonomy granting and proactive or directive control once positive grandparenting and other covariates were taken into account. Positive grandparenting also accounted for a substantial proportion of variance in prosocial behavior and overlaps conceptually with more supportive and structured distant mothering practices, which likely reduced the unique variance that could be attributed to the other positive distant mothering dimensions, even though multicollinearity diagnostics were acceptable. In this context, distant responsivity may represent the most distinctive aspect of maternal communication via mobile phones, conveying warmth and contingent emotional support, which could explain why it emerged as the only significant predictor in the multivariate models. It is also possible that the other distant mothering dimensions play a role in child adjustment primarily through indirect or child-driven pathways. For example, they may shape other family subsystems (such as coparenting or grandparenting) or be adjusted in response to children’s existing behavior. Thus, they may not emerge as strong independent predictors in our cross-sectional models.

The results suggested that children whose mothers used more responsivity in their audio or video calls also tended to be more prosocial in homes and preschools, in line with previous research on general maternal parenting (Garner, 2006; Wang et al., 2006; Pastorelli et al., 2016). Maternal responsivity can be related to child prosocial behavior in various ways. Attachment researchers argue that emotional warmth provided by mothers makes children feel secure and trusted, subsequently developing children’s sense of belonging and initiative of connecting with others (Ferreira et al., 2016). Also, mothers’ support and warmth practices can inform children of the model of sympathetic engagement, caring, and helping others (Paulus and Moore, 2012). Moreover, children may be more willing to internalize their mothers’ values regarding respect and connectedness to others when they feel the warmth from their mothers (Grusec and Goodnow, 1994). Consistent with these interpretations, our earlier qualitative study on mobile-phone-based parenting (Liang et al., 2023) showed that migrant mothers deliberately tried to use praise, comfort, and contingent responses during calls to convey warmth and support to their LBC.

However, the role of maternal distant responsivity in explaining child prosocial behavior was rather limited. This suggests that compared to other factors like genetics, temperaments, peers, preschools, and grandparenting, distant mothering may not be the most important aspect in the complex process of LBC’s early childhood socioemotional behaviors. A bidirectional influence between mothering and child behaviors (Pastorelli et al., 2016) might also complicate the examination of the individual contribution of distant mothering, as bidirectional relationships increase the difficulty in isolating the individual contribution of the independent variable which may change with the dependent variable in feedback loops. From a child-driven perspective, children’s existing prosocial tendencies and behavior problems may themselves shape how mothers behave during calls: more prosocial and easy-to-manage children may evoke warmer, more responsive interactions, whereas children who show more difficulties may elicit more controlling or even harsher responses. This interpretation is supported by our earlier qualitative study on mobile-phone-based parenting (Liang et al., 2023), in which migrant parents reported adjusting the frequency, length, and content of phone calls according to children’s mood, cooperation, and well-being, and described how some children actively requested or initiated contact with their parents. In this case, the parenting practices we observe partly reflect mothers’ responses to children’s current behavior, which can weaken or hide the direct effect of distant mothering on child adjustment in cross-sectional analyses.

On the one hand, concurrent associations between distant responsivity and child prosocial behavior may accumulate over time to demonstrate a more noticeable impact if mother–child separation is prolonged. On the other hand, as a subsystem in a family, distant mothering can interact with other parts of family dynamics (e.g., grandparenting, Acedera and Yeoh, 2021) and further indirectly impact child development and well-being.

The quality of mother-grandparent coparenting relationship was positively correlated with child prosocial behavior and negatively correlated with child difficulties. When taking demographic characteristics, positive grandparenting and distant mothering into account, the quality of mother-grandparent coparenting relationship still contributed to the variance of child prosocial behavior. In line with part of Hypothesis 3, the present findings showed that children from families with a higher quality of mother-grandparent coparenting relationship showed more prosocial behavior at home and in preschool. This result is consistent with previous research that has demonstrated positive associations between father–mother/intergenerational coparenting quality and child emotional and social competence (Teubert and Pinquart, 2010; Li and Liu, 2019; Xu et al., 2023). Family systems theory underlines the jointly-created family climate and the interdependence of family members’ feelings, values, and behaviors (Minuchin, 1985; Cox and Paley, 1997). Accordingly, children in our study might be more likely to replicate the communicative patterns between mothers and grandparents through role modeling processes to develop their prosocial tendencies if these coparenting figures frequently showed harmonious reciprocities, effective telecommunications, shared values, cooperation, and mutual support during distant interactions by electronic devices. Moreover, the prevention of situations where mothers and grandparents compete for the closeness of their (grand)child or where grandparents blame migrant mothers for the shifted childrearing duties, may promote child prosocial behavior by increasing the child’s sense of stability and security (Lau, 2017). From a cultural perspective, filial piety norms may encourage adult children to respect and defer to grandparents’ views on childrearing, which can facilitate cooperative coparenting when values are aligned but also make it difficult for migrant mothers to explicitly challenge grandparental practices when disagreements arise. Such dynamics may partly shape how mother-grandparent coparenting quality is experienced in LBC’s families.

Inconsistent with part of Hypothesis 3, the current study did not find any significant association between coparenting relationship quality and child difficulties. This may be explained by the indirect paths from mother-grandparent coparenting quality to child problem behaviors through other parenting characteristics like maternal sensitivity (Liang et al., 2021). In other words, despite the statistically insignificant link, the quality of mother-grandparent coparenting relationship may remain critical for other facets of child adjustment like child difficulties. The importance of the coparenting relationship quality in child adaptation will be gradually confirmed as future studies further include more factors of the parenting process and reveal underlying indirect mechanisms among the subsystems within a family.

Although we found that the quality of mother-grandparent coparenting relationship moderated the association between distant proactive control by mothers and child prosocial behavior, the interaction between coparenting relationship quality and proactive control was not a synergistic or enhancing pattern where the joint effects of positive mothering and good quality of coparenting relationship were greater than the total of their individual effects as previous research identified (Scrimgeour et al., 2013; Yan et al., 2021). The findings partially diverge from Hypothesis 4. Instead, our results showed the interaction had an either-or effect on the association between maternal proactive control and child prosocial behavior. This either-or effect means that the significance of the relationship between the independent variable and the dependent variable changes according to the level of the moderator (Cohen et al., 2013). In our study, children from families with both low levels of mother-grandparent coparenting relationship quality and maternal proactive control showed the lowest level of prosocial behavior, whereas there was no significant difference in levels of prosocial behavior of children at any levels of coparenting quality as long as mothers reported high levels of maternal proactive control. In other words, the risk of low quality of mother-grandparent coparenting relationship for child prosocial behavior was salient at low levels of proactive control by migrant mothers while the risk seemed to be negligible when mothers used high levels of proactive control. However, this interaction effect should be interpreted with caution, as it did not remain statistically significant after adjusting for multiple comparisons using Bonferroni and Holm-Bonferroni corrections.

Considering the multiple family-related determinants of child prosocial tendencies (Eisenberg and Mussen, 1989), we can speculate that the extent to which coparenting relationship quality and distant mothering were related to child prosocial behavior might vary in different families. Young children are selective learners in the social learning process and they can decide what and from whom to learn based on social cues like the competence of the individuals and reinforcements/rewards (Poulin-Dubois and Brosseau-Liard, 2016; Heyes, 2017). In families with medium- or high-quality coparenting relationships, children might be likely to benefit more from the modeling of cooperative and positive communications between migrant mothers and grandparents than from mother–child telecommunications. This interpretation is consistent with our earlier qualitative study (Liang et al., 2023), in which some migrant parents described how harmonious or conflictual relationships with at-home caregivers shaped the way they used mobile-phone-based guidance and discipline in everyday situations. When mothers and grandparents cannot achieve a relatively cooperative and supportive relationship and have more conflicts in childrearing, LBC might rely more on mothers’ appropriate disciplinary techniques to learn altruistic actions if their mothers show high levels of proactive control. It is possible that in the case of low quality of intergenerational coparenting relationship and high levels of maternal proactive control, the children gained more material or emotional rewards from their mothers, or they believed that their mothers were more reliable and competent figures to imitate. Mothers’ explanations, reasoning, verbal persuasion, and preaching are all proactive control behaviors, by which children can model being considerate and rational for social interactions and get information about unwanted social behaviors and the awareness of being responsible for their own behaviors (Eisenberg and Mussen, 1989; Pastorelli et al., 2016).

The study was conducted in Chuxiong Yi Autonomous Prefecture, where a substantial proportion of the population belongs to Yi and other ethnic minority groups. At the same time, Yi and Han families in this region typically live in mixed and relatively integrated communities. Although we did not collect individual-level data on participants’ ethnic identity, it is likely that our sample included both Yi and Han families. Existing work suggests that ethnic minority communities in these areas may place a strong emphasis on extended family networks, grandparental involvement in childcare, and intergenerational obligations (Chen and Zhan, 2012), which could contribute to the prominent role of grandparents and the patterns of intergenerational coparenting observed in this study. However, without direct ethnicity measures and without in-depth anthropological data on the specific communities we studied, it is difficult to disentangle the influence of Yi cultural traditions from broader rural childrearing norms in southwest China. Our discussion on cultural influences should therefore be seen as tentative.

In addition to the main aims of the current study, it is worth noting that more frequent father-child telecommunication was related to fewer child difficulties after taking into account other demographic features, positive grandparenting, and distant mothering. This is consistent with the results of previous research that found the positive influence of fathers’ involvement on internalizing and externalizing symptoms of Chinese LBC (Wu et al., 2021). Although distant fathering was not a primary focus in this substudy and was included only as a covariate, this association suggests that regular contact with migrant fathers may also act as a protective resource for LBC’s socio-emotional adjustment and warrants more systematic examination in future research.

This study was innovative, because we combined the perspectives of intergenerational coparenting and coparenting in the context of family separation and media use, which provided a new framework for LBC’s family education and intervention. Furthermore, the multi-informant reports in our study provided a more comprehensive picture of children’s behaviors across different settings (home and preschool) and the ratings of coparenting relationships by both mothers and grandparents increased validity and mitigated bias.

Several limitations in the current study could be addressed in the future. First, the modest sample size limited our ability to detect interaction effects with adequate statistical power. In additional analyses, the interaction term between maternal proactive control and coparenting relationship quality did not remain significant after adjusting p-values for multiple comparisons (Bonferroni and Holm-Bonferroni corrections). Therefore, the moderating effect reported in this study should be regarded as tentative and interpreted with caution. Second, this study focused only on positive grandparenting and the overall coparenting relationship, as measured by the short version of the CRS. Consequently, we did not differentiate between specific aspects of grandparenting (e.g., gender-biased caregiving, harsh punishment) or coparenting (e.g., coparenting support, coparenting undermining). In addition, the quality of a more general relationship between grandparents and migrant parents is also valuable to include in future research, because intergenerational solidarity, which is affected by new media and communication technologies, is important for family functioning (Taipale, 2019). Other aspects of intergenerational relationships (e.g., Grandparent’s satisfaction with the support provided by their adult children in using communication devices.) may be associated with parenting behaviors and child well-being. Third, because the number of participants in the study was limited, we could not differentiate the effects between mother-maternal grandmother, mother-paternal grandmother, mother-maternal grandfather, or mother-paternal grandfather coparenting patterns. It was thus not possible to compare group differences in the coparenting relationship-child adjustment associations among these dyads, as He et al. (2023) did. Forth, although we noticed the importance of father-child telecommunication in terms of frequency, this study did not include the quality of distant fathering, which may play a unique role in LBC’s well-being and development. Fifth, this study adopted a cross-sectional design to explore relationships between the variables of our concern, and thus cannot reveal causality and accumulative long-term effects of grandparenting, distant mothering, and coparenting relationship quality on child adjustment. Additionally, such cross-sectional design, conducted during the COVID-19 pandemic, needs replication to see whether the findings apply also in post-pandemic conditions. Sixth, child adjustment in this study was operationalized as child prosocial behavior and difficulties, and we could not test whether distant parenting practices are associated with other possible child outcomes, such as child emotional regulation, self-esteem, and well-being. Seventh, the small sample size, suboptimal sampling methods, the specific study context, and the absence of follow-up qualitative work with both parents and grandparents may limit the generalizability and depth of our findings. Eighth, we did not collect information on some potentially relevant family characteristics, such as sibling status or ethnic background. Future studies with larger samples and more extensive demographic assessments should include sibling characteristics and ethnicity to provide a more comprehensive and contextually sensitive picture of LBC’s family environments.

Given the identified unique functions of grandparenting and distant parenting for psychosocial and behavioral adjustment of LBC, separate intervention programs aiming at grandparent caregivers or migrant parents turned out effective (Chan et al., 2019; To et al., 2019) and can be recommended to stakeholders. If possible, an individualized video self-modeling approach (Regan and Howe, 2017) can be further used in families with young LBC at a higher risk of adaptive problems, to improve grandparent–child or parent-child interactions. Practitioners can use the items in the authoritative style subscale of the SRDQ and the responsivity subscale of the MPPPQ to discuss parenting practices with those caregivers.

In addition, a comprehensive family intervention program involving parents and grandparents at the same time may be a better option given the interrelatedness of (behaviors of) members within different family subsystems. Considering the risk of less child prosocial behavior in families with low-quality mother-grandparent coparenting relationships and low levels of maternal proactive control, practitioners can, if necessary, pay more attention to this group and additionally provide them with information about proactive control in distant parenting. Besides the focus on parenting practices, the integrative framework of intergenerational coparenting proposed by Bai et al. (2023) and the framework by Feinberg (2003) can inform these interventions to consider including diverse aspects that aim to establish a shared understanding among different generations regarding power dynamics, authority, and division of labor. For interventions conducted in familism-oriented societies, some researchers emphasize considering the importance of filial piety in project development (Bai et al., 2023).