Introduction

Over the past three decades, the literature in the field of developmental psychology has extensively debated the role of parenting in the socialization and adjustment of children and adolescents. However, with a paucity of combined family systems (i.e., families with three generations that is, grandparents, parents, and grandchildren living under the same roof: Che-Alford and Hamm 1999), most of the empirical studies have been conducted on nuclear family system (i.e., families consisting of only two generations that is, married parents and their unmarried children). Given that a typical family comprises the nuclear family model, the current literature lacks a focus on the grandparents’ role. However, combined family systems are not a rare phenomenon. In some cultures (particularly in Asian cultures), multigenerational or joint households are a hall mark, where grandparents, parents, and grandchildren reside under one roof and share common economic resources (Chadda and Deb 2013). According to the family systems theory, within each family, there are substructures in the form of dyadic relationships such as parent-child relationships, grandparent-grandchild relationships (i.e., relationships between members of two generations), and parent-parent or sibling-sibling relationships (i.e., relationship between members of same generation). Each relationship has its own unique characteristics and differently affects members of the family, making family ecology an important determinant of child development.

The wellbeing of individuals can be regarded as dependent upon the larger family context in which they live (Guerin and Chabot 1997). In joint families, the emotional environment in which children and adolescents are brought up is shaped not only by parents but also by grandparents (Ali and Malik 2015). The parent-child relationship is the most proximal dyadic relationship in a family and the interaction between these two entities of family systems takes place through the process of parenting. Parenting plays a crucial role in developing the competence of children and adolescents. The development and socialization process of children and adolescents is bidirectional in nature (Johnson 2010). Parents/elders convey socialization messages to children, while children/youth vary in their level of acceptance, receptivity and internalization of these messages. The internalization or perception of socialization messages results in different outcomes for children (Johnson 2010).

Parents and other family members inculcate socio-cultural values in children through different channels and practices such as reinforcement, social learning and internalization of social values (Maccoby 1992). Parents differ in the strategies they choose to nurture their children. Baumrind (1991) proposed three parenting styles i.e, authoritarian, authoritative, and permissive parenting which differ in terms of responsiveness and demandingness, and have enduring consequences on children’s developmental outcomes (Besharat et al. 2011; Buschgens et al. 2010; Hoeve et al. 2011; Jabeen et al. 2013; Kausar and Shafique 2008; Spera 2005). Authoritarian parenting is defined as a parenting style in which parents are highly directive and children are not allowed to question values. Parents demand obedience from the child. This style is characterized by parental detachment, a lack of parental warmth and use of punitive measures. Authoritative parents provide clear and firm directions for child, and are characterized by warmth, reason, flexibility and verbal give-and-take. Permissive parenting is characterized by few parental demands and the belief that child can regulate his or her own activates. These parents are non-controlling and may be warm and loving or neglectful, depending on the nature of parents (Baumrind 1991).

Empirical evidence have shown that supportive parenting styles have a significant positive effect on children’s social competence (Feldman and Wentzel 1990; Grolnick and Ryan 1989) whereas unsupportive parenting styles have negative consequences (Lamborn et al. 1991). In joint families, grandparents also share the role of socialization agents along with parents. They serve as sources of information and affection, and their emotional closeness and warmth towards their grandchildren may act as a buffer against any negative family conditions. Studies have indeed reported that grandchildren who have a close and warm relationship with their grandparents experience less psychological distress (Ruiz and Silverstein 2007). Moreover, emotional closeness with grandparents also has been shown to moderate the negative effects of family context (Attar-Schwartz et al. 2009). Emotional closeness is the degree of positive sentiments held (about grandparents) and the degree of reciprocity of those sentiments like affection, warmth and trust.

Earlier literature addressing the influence of grandparents on children has focused either on single parents or divorced families (DeLeire and Kalil 2002), but largely neglected multigenerational families. When multiple generations live under one roof, the interaction between old and young generations may have a differential role and contribute to the psychosocial adjustment of the younger generation. Additionally, recent developments in intergenerational demography (especially life expectancy) have resulted in greater opportunities for shared life years and interactions. Thus, there is a strong need to understand the influence of these interactions on the lives of individuals within families (Bengtson 2001).

In the present study, we investigated the role of emotional closeness of children with grandparents on the relationship between parenting and social competence of adolescents in joint families. We hypothesized that (1) child-reported parenting styles are a better indicator of children’s social competence compared to parents’ self-reported parenting. (2) Positive parenting styles (i.e., authoritativeness) are positively associated with social skills and negatively associated with impulsiveness, withdrawal, assertiveness and overconfidence. (3) Negative parenting styles (i.e., authoritarian and permissive) are negatively associated with social skills and positively associated with impulsiveness, withdrawal, assertiveness and overconfidence. (4) Emotional closeness to grandparents moderates the effect of parenting style on components of social competence (i.e., social skills, with impulsiveness, withdrawal, assertiveness and overconfidence).

Method

Participants

The sample included only joint families, with grandparents residing in the same home, both parents living together, and at least one of their children between 13–18 years old. The family sample included both parents and one child. Informed consent was obtained from all participants and interviews were conducted at the participants’ homes. Families were recruited from three metropolitan cities of Pakistan including twin cities Rawalpinidi and Islamabad (Capital), Lahore (Provincial Capital), and Kasur (a major urban settlement of Punjab on eastern border). A total of 130 joint families who met the inclusion criteria were approached out of which 107 (82 %) agreed to participate. Only 99 families (approximately 94 %) completed interviews. Among the 8 excluded families, either a member was not available for interview (father n = 3) or a member did not complete the interview (child n = 5). The sample for the present study consisted of 99 joint families (children n = 99; parents n = 188 (fathers n = 99; mother n = 99)). The age of fathers ranged from 32 to 65 years with mean ± sd (43.73 ± 5.99), mothers’ age ranged from 28 to 60 years with mean± sd (38.53 ± 5.70), and children’s age ranged from13 to 18 years with mean±sd (14.91 ± 2.31). Other family demographic included the number of children in the family (median = 4), the number of family members (median = 8), formal years of father’s education mean ± sd (11.80 ± 3.41), formal years of mother’s education mean ± sd (10.65 ± 3.95), formal years of adolescent’s education mean ± sd (9.80 ± 1.64), and total family income mean ± sd (60.60 ± 55.50) in 1000 PKR (Pakistan Rupees) as unit of measurement.

Procedure

The proposal of the study was approved by the institutional ethical review committee. Four interviewers including the first author of the study were trained for data collection in four different cities. The cities assigned to the interviewers were based on their residential affiliation. Joint families meeting the criteria were approached by the interviewers at their homes. Before asking for their consent to participate, all family members were well informed about the purpose of the research and they were assured for the confidentiality of their responses. The appointments for the interviews were made preferably on weekends to maximize likeliness of availability of participating family members on the same day. Families were briefed about study protocols and interviews were conducted in random order to counter systematic influence of any possible contact and sharing of information between participating family members.

Measurement

Perceptions of parenting styles were measured from all three participating family members (i.e., fathers, mothers, and children) on the three parenting styles. Children’s social competence and emotional close to grandparents was measured based on the child-reported instruments.

Parental Authority Questionnaire (PAQ)

Child-reported parenting styles of their parents on the Parental Authority Questionnaire (Buri 1991). It is a five-point Likert type scale consisting of 30 items with “1 = completely false” to “5 = completely true”. Each parenting style (i.e., authoritative, authoritarian and permissive styles) is measured by 10 items, with high scores representing high levels of the particular parenting style. Separate forms were used to measure adolescents’ opinion about fathers’ and mothers’ parenting styles. An example item to measure authoritative parenting style item includes “My parent does not allow me to question the decision that he make”. Reliability for the three subscales ranges from .77 to .92 (Buri 1991). For the present study, the translated version of the Parental Authority Questionnaire was used (Barbree 1997).

Parental Authority Questionnaire-Revised (PAQ-R)

The Parental Authority Questionnaire –Revised (Reitman et al. 2002) is a modified version of the Parental Authority Questionnaire, which poses questions to parents on their own parenting style. Similar to PAQ, PAQ-R also contains 30 items on a five-point Likert type scale with “1 = completely false” to “5 = completely true”. Each parenting style is measured by 10 items. Example items include “I tell my children what they should do, but I explain why I want them to do it” and “I do not think of myself as responsible for telling my children what to do” to measure authoritarian parenting style, and permissive parenting style respectively (Reitman et al. 2002). Reliabilities for the three subscales range from .72 to .77. For the present study, PAQ-R was translated using the back translation method (Brislin 1986).

Matson Evaluation of Social Skills for Youngsters (MESSY)

The Matson Evaluation of Social Skills for Youngsters (Matson et al. 1983) is a dimensional self-report measure to assess the social competence of children and adolescents. It consists of 62 items with a five-point Likert type scale with “1 = not at all” to “5 = always”. It measures five domains of social competence including social skills, inappropriate assertiveness, impulsiveness, withdrawal/jealousy and overconfidence. Higher scores indicating higher levels of that particular dimension. Example items include “I feel good if I help someone”, and “I think that winning is everything” for social skills, and inappropriate assertiveness respectively. For the present study, the scale was translated using the back translation method (Brislin 1986). Alpha reliability coefficients range from .60 to .88 for the five subscales (Matson et al. 1985).

Affectual Solidarity Index (ASI)

The Affectual Solidarity Index is a six-item scale posing questions to adolescents about their emotional closeness to their grandparents (Mills et al. 2001). Responses are collected on a six-point Likert type scale with “1 = rarely” to “6 = almost always. Example items include “How well do you feel your [grandparents] understand you?” The range of scores of the affectual solidarity was 5–30, with higher scores indicating higher levels of emotional closeness. For the present study, the scale was translated using the back translation method (Brislin 1986). The scale has shown good internal consistency with an alpha coefficient of .85 (Mills et al. 2001).

Data Analyses

Data were analyzed using SPSS V-21 and AMOS V-21. Cronbach’s alphas were computed to estimate the reliabilities of instruments. Confirmatory Factor Analyses (CFA) were conducted to examine construct validity of translated version of instruments. Pearson bivarite correlations were computed to examine relationships among parents’ perception and children’s perception of parenting styles and social competence of children. Hierarchical regression analyses were conducted to examine the comparative role of adolescent’s perception vs. parents’ perception in social competence of children. To test moderation by emotional closeness with grandparents, hierarchical regression analyses were revised with one additional step by incorporating the interaction between emotional closeness to grandparents and parenting styles.

Results

To confirm the factorial validation of the translated scales, confirmatory factor analyses (CFA) were conducted for the PAQ (mother, father), PAQ-R, MESSY and ASI using IBM AMOS V-21. Items were retained in their respective scale only if they fulfilled criteria (i.e., item loadings ≥ .30) (Stevens 2002), and errors were allowed to co-vary. A CFA model is considered a good fit if Cumulative Fit Index (CFI), Tuker-Lewis Fit Index (TLI), Incremental Fit Index (IFI) values are greater than .90, and Root Mean Square Error of Approximation (RMSEA) is less than .08 (Byrne 2010). As presented in Table 1, the resulting CFA models for all scales showed good fit to the data with CFI ranging from .92 to .99, TLI ranging from .91 to .99, IFI ranging from .92 to .99, and RMSEA < .05.

Table 1 Model fit Indices PAQ (father, mother), PAQ-R, ASI, and MESSY
Full size table

Preliminary analysis showed that all study variables were normally distributed, i.e., skewness < 2. Cronbach’s alpha coefficients ranged from .60 to .88 for all scales and subscales as presented in Table 2. Bivariate correlation analyses indicated that father-reported authoritarianism was negatively correlated with impulsiveness of children (r = −.21, p < .01). Contrary to the parent-report, child-reported parenting correlated with dimensions of social competence for both father and mother. According to the children’s report, fathers’ permissiveness was positively correlated with impulsiveness and over-confidence of children (r = .22, p < .05), and fathers’ authoritativeness was positively correlated with social skills (r = .30, p < .01) of children. Similarly, child-reported mother permissiveness was positively correlated with impulsiveness and over-confidence (r = .23, and r = .30 respectively; p < .05) and authoritativeness correlated positively with social skills (r = .41, p < .01) of children. Additionally, child- reported mother authoritarianism also correlated positively with impulsiveness (r = .22, p < .05) of children. No correlations appeared between parent-reported parenting and child-reported parenting (i.e., p > .05). Though no correlations appeared between father-reported parenting and mothers-reported parenting, child-reported parenting styles for both fathers and mothers were positively correlated (r ranging from .21 to .74, p < .05) except authoritativeness. Finally, child-reported emotional closeness to their grandparents were positively correlated with social skills (r = .26, p < .01), and negatively with assertiveness (r = −.21, p < .05).

Table 2 Bivariate correlation between study variables
Full size table

To investigate grandfathers’ influence on the relationship between parenting styles and the social competence of children, step wise hierarchal regression analyses were carried out. Firstly, the effects of child-reported parenting on the social competence of children were estimated by controlling for demographics (i.e., age, sex, and education) of children and parenting as reported by parents’ (both father and mother). Results presented in Tables 3 and 4 showed that child-reported parenting plays an important role in predicting the social competence of children than parent-reported parenting. Results presented in Table 3 indicate that after controlling for age, sex, and education, child-reported father authoritativeness positively predicted (β = .34, p < .05) social skills of children accounting for 7 % of the variance. Similarly, as presented in Table 4, child-reported mother authoritativeness also positively predicted (β = .43, p < .01) social skills of children explaining 18 % variance. Additionally, mother-reported authoritarianism significantly negatively predicted social skills (β = −.23, p < .05) of children explaining 5 % additional variance.

Table 3 Step wise hierarchal regression analysis for predicting social adjustment from perceived and father self-reported parenting styles (n = 194)
Full size table
Table 4 Step wise hierarchal regression analysis for predicting social adjustment from perceived and mother-reported parenting (n = 198)
Full size table

Finally, moderation by emotional closeness to grandparents was tested by repeating the regression analysis with one additional step. Results presented in Table 3 suggest that emotional closeness to grandparents moderated the effect of child-reported fathers’ authoritativeness on assertiveness (β interaction = −.37, p < .05; ΔR 2 = .13, simple slopes at high −7.29, p = .02, medium −1.68, p = .80, low 6.08, p = .25 levels of moderator), and withdrawal (β interaction = −.36, p < .05; ΔR 2 = .08, simple slopes at high −6.74, p = .02, medium −0.52, p = .39, low 5.40, p = .30 levels of moderator) of children. Additionally, as presented in Table 4, emotional closeness to grandparents also moderated the effect of child-reported mothers’ authoritativeness on overconfidence (β interaction = −.30, p < .05; ΔR 2 = .09, simple slopes at high −10.97, p = .01, medium 1.007, p = .86, low 11.89, p = .02 levels of moderator) of children. Figs. 1, 2, and 3 provide a graphical presentation of the effect of child-reported father or mother authoritativeness on child assertiveness or overconfidence in the context of different levels of child closeness to grandparents.

Fig. 1
figure 1

Moderating effect of emotional closeness with grandparents on the relationship between levels of parents authority (father) and assertiveness of children

Full size image
Fig. 2
figure 2

Moderating effect of emotional closeness with grandparents on the relationship between levels of parent authority (father) and withdrawal of children

Full size image
Fig. 3
figure 3

Moderating effect of emotional closeness with grandparents on the relationship between level of parents authority (mother) and overconfidence of children

Full size image

Discussion

The present study investigated the effects of child emotional closeness with grandparents on the differential role of child and parent-reported parenting on the social competence of children. The study was designed to incorporate both parent’s (i.e., mothers and fathers) perception of parenting styles as well as children’s perception.

Regarding the first hypothesis, our results on the differential role of child and parent-reported parenting in the determination of children’s social competence are in line with earlier literature (Barry et al. 2008) that highlighted the potential value of adolescents’ perceptions of their emotional environment. As expected, our results showed significant correlations between children’s social competence and their perception of their parents’ parenting style, in contrast to the absence of significant correlations with parent-reported parenting. The most straightforward explanation may be that children have a different view of their parents’ parenting in regard to an authoritarian parenting style when compared to the view of their parents. Either parents are underestimating their own authoritarian style of parenting or children are overestimating parents’ authoritarian parenting. In any case, whether actual or perceived, an authoritarian style of parenting appears to have a negative effect on children’s psychological development. Our results are in line with earlier research evidencing the relationship between children reports of parenting practices, and their own behavioral/social adjustment (Barry et al. 2008; Kamphaus and Frick 2005). It is rationalized that adolescence is the developmental stage when individuals can provide reliable and valid reports of parenting as well as they can be trusted for their self-report regarding their own behavioral and social functioning (Barry et al. 2008; Kamphaus and Frick 2005). The findings of the present study also support the argument that adolescents internalize the effects of parenting that they perceive which is in turn manifested in their development of social competence and adjustment.

The second and third hypotheses of the study specified relationships between parenting styles and social competence of children. Our results were in line with earlier literature suggesting that authoritarian and permissive parenting decreases social competence of children whereas authoritative parenting increases social competence of children (Baumrind 1971). These results confirmed our assumption regarding the effect of parenting styles on social competence of children, (Feldman and Wentzel 1990; Grolnick and Ryan 1989; Lamborn et al. 1991). However, recent developments in parenting literature indicated that children at risk for poorer social competence due to authoritarian and permissive parenting may benefit from other opportunities i.e., support from peer groups, schools, and families (Hart et al. 2003). The present study extends the literature by investigating role of emotional closeness to grandparents for children at risk due to unsupportive parenting styles. Our results indicated that emotional closeness to grandparents is positively associated with social skills and negatively associated with assertiveness.

Finally, the fourth hypothesis of the study stated moderating effect of emotional closeness to grandparents on parenting styles and children’s social competence. We confirmed that the grandparents’ presence and role in the home affects the relationship between children’s perceived parenting and social competence. The relationship between child-reported fathers’ level of authority and inappropriate assertiveness was moderated by emotional closeness to grandparents and showed interesting patterns under various levels of emotional closeness between children and grandparents. As presented in Fig. 1, emotional closeness to grandparents appeared to diminish the relationship between fathers’ level of authority and inappropriate assertiveness of children. In other words, children’s emotional closeness to grandparents provides a channel for the effect of fathers’ level of authority in decreasing inappropriate assertiveness of children at risk for poorer social competence. Similar patterns of relationships were observed for child reported fathers’ level of authority and child perceived withdrawn. Emotional closeness to grandparents also moderated the effect of mothers’ parenting on the social competence of children. As presented in Fig. 2, child-reported mothers’ level of authority appeared to decrease overconfidence in children under high levels of emotional closeness to grandparents. These findings showed that children at risk for poorer social competence may benefit from emotional closeness grandparents. Our results are in line with earlier literature suggesting that close interaction with grandparents is a protective factor for children at risk of maladjustment (Werner and Smith 1992).

The explanation for the current findings lies within the role of grandparents in a combined family system. Children who are emotionally close to their grandparents usually discuss their concerns with them. Grandparents in turn provide children with social and emotional support and this may fill in some of the gaps in communication between fathers and children. This finding supports the social ecological model and family systems concepts which propose that loyal and caring interactions with family members other than the immediate family (i.e., parents or siblings) are related to better adjustment of children and adolescents (Wakschlag et al. 1996). Werner and Smith (1992) reported that close interaction and communication with a grandparent is a frequently reported protective factor for children at risk for maladjustment as they provide consistent care in the times of multiple transitions. The results support the protective role of grandparents. Families could benefit from planning activates engaging both grandchildren and grandparents to enhance emotional closeness between them.

Limitations

The present study used a cross sectional method. Though we used multi-informant approach, the perception of grandparents is not covered. In addition, other intergenerational relationships and dynamics (for example parents’ siblings or cousins’ relationships) are not addressed in the present study. Cultural diversity is another limitation for generalization of results of the present study. The study shall be replicated in families with different cultural backgrounds.

Future researches should address limitations of the current study to advance our understanding of family mechanisms. A longitudinal design might result in a better understanding of grandparents’ roles in psychosocial development of children. In addition, other intergenerational relationships and dynamics (for example parents’ sibling or cousins’ relationships) could be incorporated as these may also have important influences on the development and competence of children in multigenerational families.