Bidirectional Socialization: An Actor-Partner Interdependence Model of Internet Self-Efficacy and digital Media Influence Between Parents and Children

Abstract

Media researchers have studied how parents and children influence and guide each other’s media use. Although parent and child socialization and influence are thought to be bidirectional, they are usually studied separately, with an emphasis on parental socialization, influence, and guidance of the child’s media use. In this article, we present results from a study that investigates perceived bidirectional digital media socialization between parents and children from the same household (N = 204 parent-child dyads). This study simultaneously tested parent-to-child and child-to-parent influence using the actor-partner interdependence model to examine the association between perceived Internet self-efficacy and perceived digital media influence. Although the results showed significant cross-sectional actor and partner effects for Internet self-efficacy and perceived digital media influence, these effects largely disappeared in a longitudinal setting.

Keywords

interpersonal communication topics mass media social influence structural equation modeling data analysis

Introduction

Socialization is the process by which individuals are socially guided to acquire norms, values, and practices so as to become competent members of their society and culture. Historically, socialization was seen as the outcome of a linear process of intergenerational transmission (Parsons, 1951). Even though different agents of socialization (including peers, teachers, mass media) have been acknowledged, the majority of studies have concentrated on the parent-to-child direction of influence and socialization (e.g., Grusec & Davidov, 2007; Hartup, 1979; Hill, Bush, & Roosa, 2003; Kochanska, 1993). However, for several decades, researchers have drawn attention to the importance of children as active agents in family life. This view acknowledges that children influence the socialization experiences they receive from parents (Bell, 1968; Maccoby & Martin, 1983). They interpret parental messages (Grusec & Goodnow, 1994; Kuczynski, Marshall, & Schell, 1997) and more broadly, children influence their parents’ own continuing development (Kuczynski, Pitman, Ta-Young, & Harach, 2016). New theoretical models of socialization processes are replacing unidirectional models with a perspective that emphasizes constant bidirectional interactions and influences between parents and children throughout the lifespan (Kuczynski & De Mol, 2015). In this account, parents are seen as exerting their influence on children in a relationship context where children also exert influence on parents.

Although research on media socialization is more recent, it appears to have followed a theoretical path similar to that of the traditional socialization literature. This discipline has a dominant focus on the unidirectional parent-to-child influence or parent effect model and its association with child outcomes (e.g., Mesch, 2009; Nathanson, 1999, 2002; Valcke, Bonte, De Wever, & Rots, 2010; van Den Eijnden, Spijkerman, Vermulst, van Rooij, & Engels, 2009). A child-to-parent direction of influence or child effect model is more recent (Van den Bulck, Custers, & Nelissen, 2016), and occurs most in terms of children’s digital media influence or guidance (e.g., Correa, 2014; Nelissen & Van den Bulck, 2018b). This focus on digital media is probably a result of the interest generated by the extent to which children are surrounded by digital media and their frequency of using it (Stafford, 2003). Although bidirectional models of influence have not yet been empirically tested in media research, we propose that considering how parents and children mutually influence and guide each other in the way they use digital media is the next step in this field. Therefore, the main aim of the present study is to examine a bidirectional model that incorporates the mutual influence of parents and children on digital media use in the family context.

Theoretical Framework: Bidirectional Socialization Between Parents and Children

The phenomenon of children’s influence on socialization processes in the family was identified by Bell (1968) who reinterpreted correlations between parent and child behaviors as plausibly reflecting the effects of children on parents. Subsequently, several models of bidirectional causality were developed that incorporated both parent-to-child and child-to-parent effects (Kuczynski, 2003). The most influential one was the transactional model of human development (Sameroff, 1975, 2009) which emphasized the qualitative transformations that occur when individuals and their social environments respond to each other over time. More recently, Kuczynski and colleagues (Kuczynski, 2003; Kuczynski & De Mol, 2015) developed the social relational theory as an elaboration of the transactional model with the purpose of unpacking the dynamic micro processes of social interaction that occur between parents and children during socialization. Social relational theory asserts that both parents and children are active agents, despite being unequal in power. It also asserts that influence between parents and children goes in both directions, and that the expected outcome of the coaction of parent and child influences is a continual process of qualitative change (Kuczynski, 2003). Figure 1 is a visual representation of this process. Agency in social relational theory is a multidimensional concept comprising motivation, cognition, and action. This comprehensive concept of children’s agency allows an exploration of the specific nature of the child’s active contributions to bidirectional processes (Kuczynski, 2003). Examples of such contributions include interpreting parental messages, resisting parental demands, or persuading parents to act in accord with children’s goals (Kuczynski, 2003; Saphir & Chaffee, 2002).

Figure 1.

Adaptation from the original transactional model for parent-child interactions (Kuczynski & De Mol, 2015 based on Sameroff, 1975).

Studies documenting these child effects are rare, but can be found in many research domains (Parke, 2002) including family life, value transmission, parental attitudes, health and lifestyle, and product purchases (Ambert, 2001). Despite this research stream, the unidirectional parent-to-child view on socialization often remains the default view in several fields of social research. Although the concept of bidirectionality is commonly acknowledged at a theoretical level, the implementation of it in empirical research and practice remains scarce (Collins, Maccoby, Steinberg, Hetherington, & Bornstein, 2000; Kuczynski & De Mol, 2015). The same observation can be made of the media research literature. Even though media scholars have looked at child effects and child agency, such research is rare compared with research on parental effects. For example, the literature on media guidance to date is well developed (e.g., Livingstone & Helsper, 2008; Nathanson, 1999, 2001; Nikken & Jansz, 2006, 2014; Valkenburg, Krcmar, Peeters, & Marseille, 1999), but Coyne, Padilla-Walker, Fraser, Fellows, and Day (2014) argued that most of the literature on media use in the family context focuses on parental media mediation, such as monitoring and co-using media. However, children also contribute to the media socialization of their family.

Bidirectional Socialization Between Parents and Children in Media Research

In a conceptual review, Van den Bulck and Van den Bergh (2005) proposed three ways by which children contribute to socialization processes in communication and media use. First, children can be viewed as cocreators of their own socialization. The actions and messages parents direct toward their children are interpreted by children in a way that may differ from the meanings intended by parents. For example, several researchers have acknowledged that children may not perceive parental mediation the way the parents intended it to be perceived (Haddon, 2015; Valkenburg, Piotrowski, Hermanns, & de Leeuw, 2013; Van den Bulck & Van den Bergh, 2000). Second, children use problem solving skills to socialize themselves. For example, Olesen (2000) found that children developed personal strategies for dealing with scary media content, without any intervention from their parents. Third, children can also influence their parents approach to childrearing, the process most commonly referred to as child effects (Bell, 1968).

Recognizing the latter strategy, media scholars have started to acknowledge child agency and started examining child effects, mostly within the field of digital media. The focus of child effect research on digital media is not very surprising, because most children are enthusiastic and frequent users of technology and digital media (Stafford, 2003). In many cultures recent generations of children have experienced a completely different media landscape than what their parents were exposed to when they grew up. As a result, considerable generational differences in (digital) media use and skills have been documented (e.g., Bolin, 2016; van Deursen, van Dijk, & Peters, 2011). For example, van Deursen et al. (2011) found that the younger generation was better at operational and formal Internet skills (e.g., navigating online, bookmarking websites), whereas the older generation was better at content-related Internet skills (e.g., evaluating certain information). Research has found that children have an influence on the digital media adoption and use of their parents. For example, during the years in which the Internet became a pervasive presence in many if not most households, researchers such as Selwyn (2004) and Van Rompaey, Roe, and Struys (2002) found that children influenced the adoption and use of computers by their parents. Recent studies have shown that children were also involved in actively teaching parents how to use the technology (Ito et al., 2009; Katz, 2010). Children have been shown to play a role in guiding parents’ use of the Internet (Correa, Straubhaar, Chen, & Spence, 2013), computer learning, mobile learning, and Internet learning (Correa, 2014). Although these studies concluded that children play a moderate to crucial role, other studies concluded that the role that children play is limited because several other factors influence this adoption and use process (e.g., Cáceres & Chaparro, 2019; Eynon & Helsper, 2015; Selwyn, 2004).

Although research on child influence and bidirectionality between parents and children in media research is still very limited, several predictors and outcomes have been investigated. The studies of Correa (Correa, 2014, 2015; Correa et al., 2013) examined structural factors that were associated with child influences on parental media use. These studies found that children’s influence in digital media use was predicted by lower socioeconomic status and by lower scores on authoritarian parenting. Correa and colleagues (2013) also examined the association between child-to-parent digital media influences and Internet self-efficacy. They found a marginally significant association in which more child-to-parent influence was associated with lower parental Internet self-efficacy.

Although researchers have acknowledged that children are active agents who give meaning to, and interact with parental messages, there is very little research that combines parent and child influences in media research. In an article anticipating a bidirectional perspective, Clark (2011) suggested that a new media mediation strategy, which the author called participatory learning, should be included in the media mediation literature. This strategy considers how parents and children learn from each other in their use of digital media. In a similar vein, a study of Zaman, Nouwen, Vanattenhoven, de Ferrerre, and Van Looy (2016) on parental media mediation concluded that research should start examining two-way negotiations between parents’ and children’s media use.

Objectives of the Study

The main aim of this study is to investigate parent-to-child and child-to-parent digital media influence together in a bidirectional socialization model, focusing on Internet self-efficacy and digital media influence. Self-efficacy (Bandura, 1977, 1982) is a concept that describes one’s own perception of skills and abilities. Internet self-efficacy, therefore, describes the skills that an individual believes they have to perform online tasks. Despite parents’ maturity, greater power, and their role as primary socializing agents, the context of digital media is one where children may have greater perceived expertise and consequently a greater sense of self-efficacy than parents do (Prensky, 2001; Stafford, 2003). Therefore, the first research objective of this study is to examine differences between parents and children in their perceived digital media influence and in ther Internet self-efficacy.

Research Question 1a (RQ1a): To what extent do parents and children differ in their perception of influencing each other in terms of digital media use?

Research Question 1b (RQ1b): Do parents and children differ in their perception of Internet self-efficacy?

Research has shown that both parents and children guide and influence each other’s media use (e.g., Correa et al., 2013; Valcke et al., 2010) and that self-efficacy has been found to be a strong predictor of learning (Zimmerman, 2000), motivation to learn, and behavior change (e.g., Ajzen, 2002; Bandura, 1977). Therefore, our hypothesis is that Internet self-efficacy is likely to be positively associated with experiences of guiding others in digital media use. Thus, parents and children with higher levels of Internet self-efficacy should feel more confident in their digital media use and have more motivation to guide others in their digital media use. As a second research objective, we want to examine whether Internet self-efficacy is an intrapersonal predictor of digital media influence for both parents and children, at a single point in time (i.e., cross-sectional) and across time (i.e., longitudinal). We propose the following hypotheses:

Hypothesis 1a (H1a): Children’s Internet self-efficacy is positively associated with their own perceived influence on their parents’ digital media use.

Hypothesis 1b (H1b): Children’s Internet self-efficacy at Time 1 is positively associated with their own perceived influence on their parents’ digital media use at Time 2.

Hypothesis 2a (H2a): Parents’ Internet self-efficacy is positively associated with their own perceived influence on their children’s digital media use.

Hypothesis 2b (H2b): Parents’ Internet self-efficacy at Time 1 is positively associated with their own perceived influence on their children’s digital media use at Time 2.

The final research objective of this study is to investigate bidirectionality in the interactions between parents and children, we want to investigate whether or not Internet self-efficacy is associated with interpersonal digital media influence (parent-to-child and child-to-parent). Therefore, our final hypotheses deal with the influence parents and children have on each other. Until now, research has mostly focused on either parental media guidance or children’s media influence, separately (e.g., Buijzen & Valkenburg, 2003; Correa, 2015; Correa et al., 2013; Mesch, 2006; Nelissen & Van den Bulck, 2018a, 2018b; Selwyn, 2004). This study wants to incorporate both parental and children’s influence into one model. More specifically, we want to examine to what extent children’s and parents’ perceived digital media influence of each other is related to their child and their parent’s concurrent (i.e., cross-sectional) and prior (i.e., longitudinal) Internet self-efficacy. In other words, we want to examine whether or not having higher or lower levels of self-efficacy is associated with the influence on digital media use one gets from the other part of the dyad as we hypothesize that differences in perceived influence of digital media use can probably be explained by different initial competencies that parents and children have. These competencies could affect the extent to which they believe to be influenced or to which they believe to influence each other. In this study, initial competencies were measured using Internet self-efficacy. Previous research has observed that parents give more media guidance to their children if they are less skilled (Livingstone et al., 2017; Nikken & Schols, 2015; Shin, 2013). We believe it is possible that those who score lower on Internet self-efficacy, are more likely to be perceived as in need of digital media influence from the other part of the dyad. The final hypotheses will therefore be formulated as follows:

Hypothesis 3a (H3a): Children’s Internet self-efficacy is negatively associated with how their parent perceives to influence the digital media use of their child.

Hypothesis 3b (H3b): Children’s Internet self-efficacy at Time 1 is a negative predictor of how their parent perceives to influence the digital media use of their child at Time 2.

Hypothesis 4a (H4a): Parents’ Internet self-efficacy is negatively associated with how their children perceive to influence the digital media use of their parent.

Hypothesis 4b (H4b): Parents’ Internet self-efficacy at Time 1 is a negative predictor of how their children perceive their influence on the digital media use of their parent at Time 2.

Method

Data Collection

This study was part of the FAME (Family and Media) 2 study, a larger, three-wave online panel survey among parents and children nested within the same family. The questionnaire included questions on media use within the familial context, interpersonal communication, and influence between parents and children. The current study used data from the project’s first (Time 1) and second wave (Time 2). The baseline sample of this study consisted of 204 parent-child dyads. Time 1 Data were collected in the first quarter of 2016. Research invitations addressed to the parents were distributed through a convenience sample of 11 primary and nine secondary schools in Flanders, Belgium, and through school networks. These invitations gave information about the general goal of the study, the URL to the online survey, and information about privacy and confidentiality. The only inclusion criterion was that the child of the dyad had to be between 10 and 18 years old at baseline, and that one parent and one child of the same household had to be willing and able to fill out the survey separately. Data of the second wave were collected online 6 months after Wave 1 was collected (September-October 2016). Parents and children who gave their email address for the follow-up studies were contacted online with the research invitation which included the URL to the online survey. In total, 109 dyads participated in Wave 2.

Part 1 of the online survey was addressed to the parent, whereas Part 2 was meant for the child. After the parent filled out their part, an instruction page asked the parent to let their child fill out their part separately from their parent. It was specifically stated that the child had to fill out their part alone without any interventions of the parent. The content of the questions was identical, the questions of the first part were framed from the parent’s perspective and those from the second part were framed from the child’s perspective. If there was more than one child aged between 10 and 18, the parent was asked to select the child whose birthday was first up (i.e., the next-birthday method, Salmon & Nichols, 1983). Subsequently, parents were asked to grant active consent for themselves and for their child. All children were also asked for consent. They were free to decide not to participate even if their parent had already responded. All participants were allowed to cease cooperation at any point. Dyads taking part in all of the survey or a part of it received a gift card for a large national supermarket chain. The full study protocol was reviewed and approved by the Institutional Review Board of Human Sciences of the University of Leuven.

Participants

The sample at baseline in this study consisted of 204 parent–child dyads nested in the same household in Wave 1 and 109 parent–child dyads that participated in both Wave 1 and Wave 2. These are standard sample sizes for dyadic data (Kenny, Kashy, & Cook, 2006). The parent participants were predominantly female, with 78.92% of the sample being mothers and 21.07% being fathers. Child gender was more evenly distributed, with 49.16% daughters and 50.84% sons. At baseline, the means for age were 44.13 years (SD = 5.14) for parents and 13.77 years (SD = 1.93) for children. At baseline, most children were in high school (79.46%), with the others being in secondary school (20.54%). Of the parents, 1.97% had no formal educational degree, 18.14% had a high school degree, 45.59% had the equivalent of a community college degree, and 34.31% had a university degree. Most parents reported working full-time (66.66%) at baseline, 18.14% worked part-time, and 7% identified themselves as a stay-at-home parent.

Regarding family composition at baseline, 97.54% of the parents in the study were the biological parent of the child filling out the survey. Fifty-nine percent of the parents lived together with their children and spouse. Another 24.88% lived together with their children and a partner they were not married to. Sixteen percent of the parents lived together with their children (full-time or part-time) but without a partner.

Attrition Analyses

Because this study had 204 dyads participating in Wave 1, but only 109 in Wave 2 (53%), we controlled for attrition bias. Specifically, we conducted a MANOVA to test if the vectors of Internet self-efficacy and digital media influence are sampled from the same distribution in Wave 1 and Wave 2. Based on Pillai’s Trace statistic, V = 0.01, F(4, 165)= 0.218, p > .05, $η_{p}^{2} = . 01$ , we found no significant differences between the respondents who participated in only one wave and those who participated in two waves in terms of their reported Internet self-efficacy and digital media influence.

Measures

Sociodemographics

Parents and children were asked to describe their gender, “male/boy” (= 0) and “female/girl” (= 1). Date of birth was recoded into a discrete variable with the current age of both the parent and their child. The educational level of the parent was also registered, ranging from “no degree” (= 1) to “university degree” (= 4).

Internet self-efficacy

The FAME2 Internet self-efficacy scale was based on the scale of Correa and colleagues (2013). The scale contained seven items about how confident the respondents felt undertaking certain online activities: “I feel confident to . . .” (upload photo’s, video’s, files; block spam or unwanted content; adjust my privacy settings on a website; bookmark a website or add a website to my list of favorites; compare different sites with verify the accuracy of information; create and manage my own personal profile on a social network site; create and manage my own personal website). Response categories were presented on a 5-point Likert-type scale ranging from “totally disagree” (= 1) to “totally agree” (= 5). For the parents’ Internet self-efficacy items, a principal axis factor analysis with oblique rotation generated a single factor that explained 62.44% of the variance (EV = 4.37). Factor loadings ranged from 0.56 to 0.88, and scale reliability was high (Cronbach’s α = .90). For the children’s Internet self-efficacy, the analyses generated nearly identical results: one factor (EV = 4.36) explaining 62.32% of the variance, with factor loadings from 0.66 to 0.87 and high reliability (Cronbach’s α = .90).

Parent-to-child digital media influence

Adapted from Correa’s (2014) measure of perceived digital influence on digital learning, the FAME2 survey queried parent-to-child digital media influence with direct questions about perceived influence on digital media use in general: “In general, do you think you have an influence on your child’s . . . ?” followed by several types of media use: computer use, Internet use, tablet use, Smartphone, and app use. We used the same 5-point Likert-type scale ranging from “(almost) never” (= 1) to “(almost) always” (= 5). A principal axis factor analysis with oblique rotation generated one factor explaining 74.11% of the variance (EV = 2.96), with factor loadings from 0.84 to 0.89. Scale reliability was high (Cronbach’s α = .88). For the descriptive analyses, we created an index of this influence by averaging the items into one score ranging from 1 to 5, where higher scores indicate higher levels of perceived influence.

Child-to-parent influences digital media influence

For child-to-parent digital media influence, children were asked similar questions about their perceived influence on the digital media use of their parents in general: “In general, do you think you have an influence on your parent’s . . . ?” This was again measured for computer use, Internet use, tablet use, Smartphone, and app use on the same 5-point Likert-type scale. The principal axis factor analysis with oblique rotation yielded one factor explaining 72.38% of the variance (EV = 2.90), with factor loadings from 0.71 to 0.90 and high scale reliability (Cronbach’s α = .87). For the descriptive analyses, we made an index of this influence by averaging the items into one score ranging from 1 to 5, where higher scores indicate higher levels of perceived influence.

Statistical Analyses

To test bidirectional influences simultaneously, this study uses dyadic data and applies a statistical analysis strategy drawn from what is known as the actor-partner interdependence model (Cook & Kenny, 2005; Kenny et al., 2006). The term dyadic data refers to situations where respondents are paired and the individuals within those pairs are nested in the same high-level entity. In the current study, we examine parents and children within one family, meaning that every individual in the sample was linked to one other individual in the sample. Most widely used inferential statistical techniques such as ordinary least squares regression assume independence of responses. For dyadic data, however, this assumption is untenable: because parents and children come from the same households they are likely to be more similar to each other than two non-related individuals (Kenny et al., 2006). This is called nonindependence, which is the core of quantitative dyadic data analysis and which can be tested by calculating a standard Pearson correlation coefficient (Kenny et al., 2006).

To test whether the reports of digital media influence and self-efficacy of parents and children differ (RQ1a and RQ1b), paired samples t tests are calculated to compare the answers of the parents with the children. To test whether self-efficacy is associated with digital media influence within a person (H1a, H1b, H2a, and H2b), and to examine whether it is associated between persons (H3a, H3b, H4a, and H4b), this study makes use of the actor-partner interdependence model (Cook & Kenny, 2005). This model is widely used in the psychology and family interaction literature to study bidirectional interaction and influence processes (Cook & Kenny, 2005; Kenny et al., 2006). Figure 2 is a visual presentation of the model. The model takes mutual interdependence into account and was created for standard dyadic designs. It allows researchers to determine how outcomes are influenced by other members of the dyad, by incorporating both intrapersonal (i.e., actor effects, in this study H1 and H2 deal with this) and interpersonal (i.e., partner effects, in this study H3 and H4 deal with this) into one model (Kenny et al., 2006). Actor effects (A) are effects of individual characteristics on outcomes within the same individual (i.e., a typical regression result). Partner effects (P), in contrast, are effects of individuals’ predictor variables on their partner’s outcome variable scores (Kenny et al., 2006). Several composite measures were used in this model: Internet self-efficacy for parent (seven items) and children (seven items), parent-to-child (four items) and child-to-parent (four items) digital media influence. The cross-sectional model was tested on the responses of Time 1, controlling for possible confounders such as child and parent gender and age, and the parent’s educational level. The longitudinal model tested relationships between Internet self-efficacy at Time 1 and perceived digital media influence at Time 2, whereas controlling for perceived digital media influence at Time 1.

Figure 2.

Visual representation of the actor-partner interdependence model (Cook & Kenny, 2005) used in this study among parents and their child (a = actor effects, p = partner effects, single-headed arrows present predictive paths, double-headed arrows present correlated variables).

All data analyses were conducted with the Statistical Package for Social Sciences (version 24, SPSS Inc., Chicago, IL, US) and AMOS (version 24, SPSS Inc., Chicago, IL, US). Descriptive data, correlation analyses, paired samples t tests, factor analyses, and reliability analyses were all done in SPSS, while the actor-partner interdependence model required estimating structural equations using the maximum likelihood method in AMOS. For the model itself, latent variables were used. The chi-square to degrees of freedom ratio (χ²/df, value between 1 and 3), the root mean square error of approximation (RMSEA, value ≤ .08), and the comparative fit index (CFI, value ≥ .95) were used to assess the fit of all models (Byrne, 2010). To investigate whether the partner effects of parent and child were significantly different, we also conducted multiple-group comparison tests. This was done by comparing a model in which the partner paths were allowed to vary across groups (i.e., the unconstrained model) with a model in which the partner paths were fixed to be equal across groups (i.e., the constrained model).

Results

Descriptive Statistics and Correlations

Table 1 summarizes the descriptive data of the constructs used in this study. Overall, the index of perceived parent-to-child digital media influence was higher than the index of perceived child-to-parent digital media influence of the children on the parents (M_parentT1 = 3.21, SD_parentT1 = 0.87; M_childT1 = 2.42, SD_childT1 = 1.10, M_parentT2 = 3.07, SD_parentT2 = 0.89; M_childT2 = 2.20, SD_childT2 = 0.95). Paired samples t tests showed that this difference was statistically significant, Time 1: t(174)= 6.82, p < .001, Cohen’s d = 0.52, Time 2: t(91)= 5.61, p < .001, Cohen’s d = 0.58. In answer to RQ1a, parents reported a higher perceived influence on their children’s digital media use than vice versa. These results need to be understood contextually: although the separate items of this index indicated that both parents and children agreed that their relative influence depended on the type of medium. For both Time 1 and Time 2, parents perceived to have the greatest influence on their children’s use of the Internet followed by (in order of importance) computer use, tablet use, and Smartphone use. Children perceived that they have the greatest influence on their parents’ use of Smartphones, followed by (in order of importance), computer use, Internet use, and tablet use. In Table 1, the separate item scores of this index are presented.

Table 1.

Means and Standard Deviations of Variables of Interest in This Study Measured at Time 1 and Time 2.

	M	SD	Range	N (dyads)
Parent-child influence computer (T1)	3.38	0.95	1-5	204
Parent-child influence computer (T2)	3.13	0.97	1-5	107
Parent-child influence Internet (T1)	3.39	0.87	1-5	203
Parent-child influence Internet (T2)	3.18	0.95	1-5	107
Parent-child influence tablet (T1)	3.09	1.12	1-5	203
Parent-child influence tablet (T2)	2.99	1.07	1-5	105
Parent-child influence Smartphone (T1)	2.98	1.13	1-5	203
Parent-child influence Smartphone (T2)	2.89	1.03	1-5	107
Child-parent influence computer (T1)	2.42	1.26	1-5	178
Child-parent influence computer (T2)	2.14	1.06	1-5	95
Child-parent influence Internet (T1)	2.38	1.26	1-5	178
Child-parent influence Internet (T2)	2.17	1.16	1-5	95
Child-parent influence tablet (T1)	2.29	1.34	1-5	176
Child-parent influence tablet (T2)	2.12	1.15	1-5	93
Child-parent influence Smartphone (T1)	2.61	1.34	1-5	178
Child-parent influence Smartphone (T2)	2.35	1.17	1-5	95
Parent-child influence digital media (T1)	3.21	.87	1-5	201
Parent-child influence digital media (T2)	3.07	.89	1-5	105
Child-parent influence digital media (T1)	2.42	1.10	1-5	176
Child-parent influence digital media (T2)	2.20	.95	1-5	93
Internet self-efficacy parent (T1)	3.24	1.02	1-5	203
Internet self-efficacy parent (T2)	3.30	0.95	1-5	108
Internet self-efficacy child (T1)	3.44	1.07	1-5	175
Internet self-efficacy child (T2)	3.53	0.98	1-5	93
Age parent (T1)	44.13	5.14	34-54	72
Age child (T1)	13.77	1.93	10-18	173
Educational degree (T1)	4.05	0.89	1-5	173

Note. T1 = Time 1, T2 = Time 2, N_Time1 = 204 dyads, N_Time2 = 109.

On the measure of Internet self-efficacy at Time 1 and 2 (RQ1b), children reported a slightly higher level of Internet self-efficacy than their parents did (M_parentT1 = 3.24, SD_parentT1 = 1.02; M_childT1 = 3.45, SD_childT1 = 1.08; M_parentT2 = 3.30, SD_parentT2 = 0.94; M_childT2 = 3.52, SD_childT2 = 0.98). The differences between parent and child were significant for Time 1, t(174) = −2.32, p < .05, Cohen’s d = −0.18. This was not the case at Time 2, t(92) = −1.75, p = .083, Cohen’s d = −0.18.

The parent-to-child and the child-to-parent digital media influence measures were significantly and negatively correlated (r_Time1 = −0.26, p < .05). This indicates that when a parent perceived to have more influence on the digital media use of the child, the child perceived to have less influence on the digital media use of the parent. Alternatively, when a child perceived to have more influence on the digital media use of the parent, the parent perceived to have less influence on the digital media use of the child. Table 2 presents the Pearson correlation coefficients of the variables in this study, with the nonindependence values in bold. Internet self-efficacy was positively correlated with parent-to-child digital media influence, indicating that parents who perceived to have better Internet competencies themselves, also perceived themselves as more influential and vice versa. A similar pattern held for children: Children’s perceived Internet self-efficacy was positively correlated with their own perceived digital media influence.

Table 2.

Pearson Correlation Matrix of Self-Reported Influence and Internet Self-Efficacy at Time 1 and Time 2.

	1	2	3	4	5	6	7	8
1. Parent-child digital media influence T1	—
2. Child-parent digital media influence T1	−.23**	—
3. Parent Internet self-efficacy T1	.36***	−.26*	—
4. Child Internet self-efficacy T1	−.26***	.36***	.17	—
5. Parent-child digital media influence T2	.53***	−.13	.29**	−.30**	—
6. Child-parent digital media influence T2	−.20	.43***	−.15	.20	−.26*	—
7. Parent Internet self-efficacy T2	.29**	−.23*	.74***	−.03	.41***	−.21*	—
8. Child Internet self-efficacy T2	−.18	.28**	.15	.77***	−.22*	.25*	−.00	—
M	3.21	2.42	3.24	3.44	3.07	2.20	3.30	3.53
SD	0.87	1.10	1.02	1.07	0.89	0.95	0.95	0.98

Note. N = 109 parent-child dyads. Correlations are zero-order correlations. Bold values represent the nonindependence between the parent and child nested in the same family.

Correlation is significant at the .05 level (two-tailed). **Correlation is significant at the .01 level (two-tailed). ***Correlation is significant at the .001 level (two-tailed).

In contrast, children’s Internet self-efficacy was negatively related to parent-to-child digital media influence, indicating that when children’s own perceived Internet self-efficacy was higher, the parental influence they received was lower, and vice versa. The same negative correlation was found for parents’ Internet self-efficacy and child-to-parent digital influence.

Testing a Model of Bidirectional Influences: The Application of the Actor-Partner Interdependence Model

We found significant intrapersonal and interpersonal correlations (see Table 2). However, to test the relationship between Internet self-efficacy and perceived interpersonal digital media influence in a single model, the actor-partner interdependence model was utilized. We applied this model of Internet self-efficacy and perceived digital media influence both cross-sectionally (H1a, H2a, H3a, and H4a) and longitudinally (H1b, H2b, H3b, and H4b).

Cross-sectional results

The cross-sectional model included Internet self-efficacy of parent and child, perceived digital media influence of the other part of the dyad, and several control variables such as gender of the parent, gender of the child, age of the parent, age of the child, and educational attainment of the parent. This model is presented in Figure 3. The model produced a relatively good fit (χ²(293) = 536.22, p < .001, CFI = 0.91, and RMSEA = 0.06 (90% confidence interval [CI] = [0.06, 0.07]), χ²/df = 1.83). All pathways in the cross-sectional model were statistically significant, indicating that support was found for intrapersonal (H1a and H2a) and interpersonal influences (H3a and H4a). Thus, in terms of intrapersonal relationships and in support of H1a, children who perceived that they had a higher Internet self-efficacy, perceived to guide the digital media use of their parents more (B = 0.29, p < .01). In support of H2b, also parents with higher perceived Internet self-efficacy, reported to have to guide the digital media use of their children more (B = 0.22, p < .001). Looking at the interpersonal relationships, or bidirectional, partner relationships, H3a and H4a were supported. Children with higher levels of perceived Internet self-efficacy had parents who perceived to guide their child’s digital media use less (B = −0.22, p < .05). The same pattern was found for the parents of our sample: parents with higher levels of perceived Internet self-efficacy had children who perceived to guide their parents’ digital media use less (B = −0.29, p < .001). The control variables had little predictive power, only the educational degree of the parent positively predicted parent-to-child digital media influence (B = 0.16, p < .05). This model explained 20% of the variance in perceived child-to-parent digital media influence (R² = .20) and 28% of the variance of perceived parent-to-child digital media influence (R² = .28).

Figure 3.

Actor-partner interdependence model of parents’ and children’s Internet self-efficacy and perceived digital media guidance in a cross-sectional model (based on answers Time 1).

To test whether the parent-to-child partner effect differed from the child-to-parent partner effect, we conducted a multiple group comparison test. To do this, we compared a model in which all path coefficients in the actor-partner interdependence model were free to vary with a model in which the pathways for the parent-to-child and child-to-parent effects were constrained to be equal. The results suggested that the model fit did not significantly differ between the unconstrained and constrained models, Δχ²(1) = 1.604, p > .05. In other words, the magnitude of parental effects on children and children’s effects on parents did not differ significantly.

Longitudinal results

Following this cross-sectional analysis, the actor-partner interdependence model was tested with Internet self-efficacy at Time 1 and perceived digital media influence at Time 2. Control variables in this model were perceived digital media influence at Time 1 for both parent and child and educational degree of the parent. This model is presented in Figure 4. This longitudinal model did not generate a good model fit, χ²(416) = 756.24, p < .001, CFI = 0.84, RMSEA = 0.09 (90% CI = [0.08, 0.10]), χ² / df = 1.82. In contrast to the cross-sectional model, none of the actor and partner pathways in the longitudinal model were significant when digital media influence of Time 1 was included as control. Therefore, we did not find support for interpersonal effects between Internet self-efficacy and digital media influence for both child (H1b) and parent (H2b), nor did we find support for intrapersonal effects over time, going from child-to-parent (H3b) or from parent-to-child (H4b). Digital media influence at Time 2 was significantly predicted by digital media influence at Time 1, for both children (B = 0.40, p < .001) and for parents (B = 0.45, p < .001). This model explained 33% of the variance in children’s perceived digital media influence (R² = .33) and 34% in the variance of parents’ perceived digital media influence (R² = .34).

Figure 4.

Actor-partner interdependence model of parents’ and children’s Internet self-efficacy and perceived digital media influence in a longitudinal model (based on answers Time 1 and Time 2).

Discussion

This study built on previous research documenting children’s influence on parental media use and vice versa by examining bidirectional influence processes between parents and their children. Several fields of research have already explored the idea of bidirectional socialization in the family, but in communication and media research it has rarely been investigated. A vast number of studies within this field have focused on guidance and influence of parents on their children’s digital and traditional media use (e.g., Nikken & Schols, 2015) or, to a lesser extent, on the influence of children on their parents’ media use and learning, separately (e.g., Correa, 2014; Correa et al., 2013; Katz, 2010; Nelissen & Van den Bulck, 2018b; Van Rompaey et al., 2002). However, to our knowledge, this has never been incorporated into one bidirectional, quantitative model. Hence, the main aim of this study was to investigate the association between Internet self-efficacy and digital media influence true a bidirectional model. We attempted to address several critical gaps in the media research literature by (1) investigating both parents’ and children’s perspective of guiding each other’s digital media use, (2) in one dyadic model, and (3) over time.

Discussion of the Descriptive Findings

In reference to research objective 1, the descriptive findings of this study showed two patterns. First, although children reported having an influence on their parents’ digital media use, the parents reported higher levels of influence on their children’s media use (RQ1a). This difference in perceived influence was contingent on media type. Second, children reported higher levels of Internet self-efficacy than their parents did (RQ1b). The parent’s and children’s Internet self-efficacy marginally increased from Time 1 to Time 2. Perceived digital media influence decreased over time. It makes sense to find that mastery increases with use (i.e., over time) in the case of self-efficacy. Regarding perceived influence, as children grow older, parent-child dyads tend to interact less about media use as this increasingly becomes part of children’s personal sphere of action (Smetana, 1995).

Parents believe to influence their children more in the area of digital media while they report lower levels of Internet self-efficacy. The question, then, is whether parents underestimate their own Internet literacy or overestimate their parental influence. Social desirability might play a crucial role in this regard (Livingstone, Mascheroni, Dreier, Chaudron, & Lagae, 2015). The idea that parents have or should have more influence than children is part of the social construction of parenthood (Kuczynski, Lollis, & Koguchi, 2003). For example, during their in-depth interviews, De Mol and Buysse (2008) reported that some parents became angry when talking about child influences because they confused influence with power and insisted that their children were not in charge. Similarly, some parents in our study may have interpreted “influence” as the ability to control or restrict children’s use of media rather than seeing it as guidance or as transmission of expertise.

The current study used dyadic data and found a negative and significant correlation between perceived parent-to-child and child-to-parent digital media influence. A significant (positive or negative) correlation between the answers of parents and children from the same dyad indicates interdependence between the members of that dyad. Negative nonindependence in this regard implies that when children influence the media behavior of their parent more, the parents influence their children less and vice versa. In the dyadic literature a potential explanation for negative interdependence has been defined as division of labor (Kenny et al., 2006). It suggests that an increase in influence from one side of the dyad to the other, is made possible by (or creates) a decrease in influence in the reverse relationship.

Discussion of the Actor-Partner Interdependence Model

For our cross-sectional actor-partner interdependence model, all hypotheses found support, but the longitudinal model did not have a good fit and the hypotheses concerning this model were not supported. In the cross-sectional actor-partner interdependence model, all pathways were significant. With regard to research objective 2 on intrapersonal associations (i.e., actor effects), we found a positive association between Internet self-efficacy and perceived digital media influence for both child and parent, confirming H1a and H2a. In other words, the higher a parent or a child perceived their Internet self-efficacy to be, the more they perceived to guide the digital media use of their child or parent. This resonates with research showing that self-efficacy is being a good predictor of learning, motivation to learn, and changing behavior (e.g., Ajzen, 2002; Bandura, 1977; Zimmerman, 2000).

With regard to research objective 3 on interpersonal associations (i.e., partner effects), the structural equation modeling showed negative associations between Internet self-efficacy and digital media influence, confirming H3a and H4a. This means that parents and children who were perceived to have lower levels of Internet self-efficacy, had dyadic partners who were perceived to have more digital media influence. In the dyadic literature, the existence of negative partner effects and positive actor effects (or vice versa) is known as a contrast patterns (Kenny & Ledermann, 2010). For parental media influence, this reproduces earlier research: Parents guided the media behavior of their children more if their children were less skilled (Livingstone et al., 2017; Nikken & Schols, 2015; Shin, 2013). The current study showed that this association also holds for children’s media influence: If their parent reported lower levels of Internet self-efficacy, the child perceived to influence the parent’s digital media use more. In other words, this contrast pattern suggests that children/parents who have a lower level of Internet self-efficacy, receive more digital media influence from the other part of the dyad and vice versa. This is consistent with several recent studies who found that children with higher media skills received less media guidance and were less influenced by their parent (Livingstone et al., 2017; Nikken & Schols, 2015; Shin, 2013). Our study found that this was also the case for parents with higher levels of skills. Interestingly, no significant difference was found between the parent and child effect.

In the longitudinal model, none of the partner and actor effects were significant, therefore we had to reject H1b, H2b, H3b, and H4b. The model did not fit the data: although the χ² and RMSEA indices were acceptable, the CFI was insufficient. A low CFI may indicate that the correlations between the variables in the model are low, which was consistent with the nonsignificant pathways in our model. Perceived digital media influence was predicted by earlier perceived digital media influence. The results of our cross-sectional and longitudinal model show that Internet self-efficacy is related to interpersonal influence of parent and child but this self-efficacy influence is not related to change in interpersonal digital media influence. This could be the result of methodological issues. It is possible that the 6 month time lag is too long and that self-efficacy of parent and child affects digital media influence contemporaneously or “on the spot” but not over time. In other words, the extent to which the other partner of the dyad influences your digital media use is related to how you rate your Internet self-efficacy at that point in time. It is also possible that a 6-month time lag is too short to produce changes large enough to be meaningful in self-reported data. Finally, given that the cross-sectional model was tested on 204 parent-child dyads, although the longitudinal model included only the 109 dyads that participated in both waves, it has to be noted that the longitudinal model had considerably less statistical power.

In summary, this study found two indications of bidirectionality between parents and children in media use. First, self-reported indications of bidirectional influences, as parents and children both reported to influence each other’s digital media use. Second, statistical indications of bidirectional influences as higher Internet self-efficacy of the parent was associated with lower child-to-parent digital media influence and vice versa (i.e., significant partner effects). These results have societal, theoretical, and methodological implications. First, we believe that the patterns found suggest that each part of the dyad will take a certain role in this process and that parents and children complement each other. Although children might teach and help their parents more with how to use certain digital media or teach operational skills, parents seem to influence their children more in terms of content, media rules, or standards of conduct. These patterns would explain why we found that parents felt to have the largest influence on their children’s Internet use (possibly focused on regulation of use and content) while children felt to have the largest influence on parents’ Smartphone use (possibly functional influence). Earlier studies (such as Van Deursen et al., 2011) found generational differences in Internet skills, which in our study was found to be associated with digital media influence. If our explanation holds, our model suggests a complementary, iterative process of parents and children influencing each other in their (digital) media use. A second and more theoretical implication is that as this study indicates that the bidirectional socialization perspective exists in media use, this bidirectional framework could be extrapolated to the study of media selection, broader media use, and media effects. In other words, future research could examine how parents and children influence each other in their media selection and media use. Moreover, as they might influence each other’s media selection and content, they might indirectly also influence the media effects that they are experiencing. We believe it is important to rethink and reinterpret the position of the child in interactions about media use and to incorporate bidirectional socialization in future theoretical models of media research in the familial context. Thirth, the results of this study also indicate that in practice, media researchers (when providing guidelines for parents) should be aware that children are active agents who interpret parental influence and guidance and who exert influence and guidance themselves. In media research, we often rely on individual approaches of constructs that inherently contain a bidirectional aspect, think about interactions, media guidance, and family media use. Starting from the individual is a common practice in social sciences (Kenny et al., 2006). Fourth, and a methodological implication, this study could have implications for reconsidering longitudinal survey research with a long time intervals when examining family interactions and influences. Our cross-sectional version of the actor-partner interdependence model found significant actor and partner effects, whereas the longitudinal model did not find this. Family dynamics and interpersonal influences can be volatile, which makes that when we use these dynamics as an outcome variable, it is possible to find different results at different times. This is information that future researchers could keep in mind when setting up new studies to investigate parent-child influences and media use.

Limitations and Suggestions for Future Research

Several limitations should be taken into account when interpreting these results. First, it is hard to ascertain how much under- and overreporting occurred in this study and, therefore, to what extent relationships may have been under- or overestimated. Perceived competences are a sensitive topic to query as it is likely to induce social desirable answers of competent skills. In addition, given that some influence on media use is subtle and/or indirect it may be difficult to form an accurate image of one’s own influence and the influence of the partner in the dyad. As mentioned earlier, parenting norms may influence the willingness to give an accurate estimate of certain behaviors. The measure of digital media influence was one of perceived influence, which could be an indication of influence and guidance but this is not necessarily the same thing. Influence is a broader umbrella term, and guidance is only one form of influence. There are potentially different meanings that may be attributed to the word “influence”: does it refer to guiding each other, to the transmission of expertise, or is it about control and restricting each other’s behavior? A lot of conceptual work remains to be done in this domain. In addition, future research should examine whether this bidirectional model holds for different types of mediation (i.e., active mediation, restrictive mediation, social co-use). Moreover, perceived digital media influence was measured for different media types and for Internet use, whereas self-efficacy focused only on Internet use in general and did not distinguish different media types. This is also something to take into account for future research. Furthermore, we used both hardware devices (computer/laptop, tablet, Smartphone) as content (Internet) together in our digital media influence measure, but future research should further examine how to measure influence of functional media use and the content of media use separately. Future studies would clearly benefit from better defining “influence” as a social scientific concept and using that definition to develop measures and participant instructions.

Future researchers could build further on the extensive parental media mediation literature for more detailed definitions and to develop appropriate measures. In addition, although this study investigated parent-child interactions, influences, and bidirectionality, we are aware that as these children grow older, peer influences also gain importance. It would be interesting for future research to incorporate several sources of influence on the individual (Bronfenbrenner, 1986). Second, this study used a sample of parent-child dyads based on a convenience sampling strategy which resulted in an overrepresentation of mothers and higher educated individuals in the parents’ subsample. Therefore, these results do not allow us to generalize our conclusions to the entire population of Belgian parents and children or beyond. However, the main goal of this study was not to generalize these results for the population, but to examine bidirectionality between parents and their children, the results may still only be valid for this particular subgroup. Finally, it would be interesting to explore which subgroups show the strongest relationships. For example, Correa (2014) found that mothers were more influenced by their children than fathers were. It could be interesting to investigate gender differences in both parents and children and see whether this moderates the association for any of the dyadic combinations. In addition, research in the area of bidirectional actor-partner relationships could really benefit from more qualitative approaches such as in-depth interviews, diary studies, and observational studies to understand how exactly individuals exert influence over their dyadic partners and what kind of implicit and explicit communication accompanies such behavior.

To conclude, media research to date has mostly investigated parent-to-child media influence, but this is only one side of a two-sided story. Most parental responses are, by definition, a response to the behavior of a specific child, and are thus already child-driven (e.g., de Haan, Deković, & Prinzie, 2012; Kerr, Stattin, & Özdemir, 2012). In addition, children also influence their parents’ media behavior (e.g., Correa et al., 2013; Nelissen & Van den Bulck, 2018b). Therefore, this study included both parent-to-child influence and child-to-parent influence into one bidirectional actor-partner interdependence model. Based on the results of our study, we believe it is important to rethink and reinterpret the position of the child in interactions about media use and to incorporate bidirectional socialization in future theoretical models of media research in the familial context.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This research was funded by the Research Foundation–Flanders [grant number G062414N].

ORCID iDs

S. Nelissen

J. Van Den Bulck

Author Biographies

S. Nelissen, PhD, is a research fellow and research advisor at the University of Leuven. Her PhD-research focused on family dynamics in media research.

L. Kuczynski, PhD, is a professor at the University of Guelph. He is interested in parent-child dynamics and bidirectional influences in socialization.

L. Coenen, PhD, is a postdoctoral researcher at the University of Leuven. He is interested in the conceptual, methodological, and statistical foundations of media effects studies.

J. Van den Bulck, PhD, DSc, is a professor at the University of Michigan. He is interested in the effects of the media on topics regarding violence, health, and sleep.

References

Ajzen

(2002). Perceived behavioral control, self-efficacy, locus of control, and the theory of planned behavior. Journal of Applied Social Psychology, 32, 665-683. doi:10.1111/j.1559- 1816.2002.tb00236.x

Ambert

(2001). The effect of children on parents (2nd ed.). New York, NY: Haworth Press.

Bandura

(1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215. doi:10.1016/0146-6402(78)90002-4

Bandura

(1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122-147. doi:10.1037/0003-066X.37.2.122

Bell

R. Q.

(1968). A reinterpretation of the direction of effects in studies of socialization. Psychological Review, 75, 81-95. doi:10.1037/h0025583

Bolin

(2016). Media generations: Experience, identity and mediatised social change. New York, NY: Routledge.

Bronfenbrenner

(1986). Ecology of the family as a context for human development: Research perspectives. Developmental Psychology, 22, 723-742. doi:10.1037/0012-1649.22.6.723

Buijzen

Valkenburg

(2003). The unintended effects of television advertising: A parent-child survey. Communication Research, 30, 483-503. doi:10.1177/0093650203256361

Byrne

B. M.

(2010). Structural equation modeling with AMOS: Basic concepts, applications and programming. Mahwah, NJ: Lawrence Erlbaum.

10.

Cáceres

R. B.

Chaparro

A. C.

(2019). Age for learning, age for teaching: The role of inter- generational, intra-household learning in Internet use by older adults in Latin America. Information, Communication & Society, 22, 250-266. doi:10.1080/1369118X.2017.1371785

11.

Clark

L. S.

(2011). Parental mediation theory for the digital age. Communication Theory, 21, 323-343. doi:10.1111/j.1468-2885.2011.01391.x

12.

Collins

W. A.

Maccoby

E. E.

Steinberg

Hetherington

E. M.

Bornstein

M. H.

(2000). Contemporary research on parenting: The case for nature and nurture. American Psychologist, 55, 218. doi:10.1037/0003-066X.55.2.218

13.

Cook

W. L.

Kenny

D. A.

(2005). The actor-partner interdependence model: A model of bidirectional effects in developmental studies. International Journal of Behavioral Development, 29, 101-109. doi:10.1080/01650250444000405

14.

Correa

(2014). Bottom-up technology transmission within families: Exploring how youths influence their parents’ digital media use with dyadic data. Journal of Communication, 64, 103-124. doi:10.1111/jcom.12067

15.

Correa

(2015). The power of youth: How the bottom-up technology transmission from children to parents is related to digital (in) equality. International Journal of Communication, 9, 1163-1186. doi:1932–8036/20150005

16.

Correa

Straubhaar

J. D.

Chen

Spence

(2013). Brokering new technologies: The role of children in their parents’ usage of the internet. New Media & Society, 17, 483-500. doi:10.1177/1461444813506975

17.

Coyne

S. M.

Padilla-Walker

L. M.

Fraser

A. M.

Fellows

Day

R. D.

(2014). “Media time= family time” positive media use in families with adolescents. Journal of Adolescent Research, 29, 663-688. doi:10.1177/0743558414538316

18.

de Haan

A. D.

Deković

Prinzie

. (2012). Longitudinal impact of parental and adolescent personality on parenting. Journal of Personality and Social Psychology, 102, 189-199. doi:10.1037/a0025254

19.

De Mol

Buysse

. (2008). The phenomenology of children’s influence on parents. Journal of Family Therapy, 30, 163-193. doi:10.1111/j.1467-6427.2008.00424.x

20.

Eynon

Helsper

(2015). Family dynamics and Internet use in Britain: What role do children play in adults’ engagement with the Internet? Information, Communication & Society, 18, 156-171. doi:10.1080/1369118X.2014.942344

21.

Grusec

J. E.

Davidov

(2007). Socialization in the family: The role of parents. In Grusec

J. E.

Hastings

P. D.

(Eds.), Handbook of socialization: Theory and research (pp. 284-308). New York, NY: Guilford Press.

22.

Grusec

J. E.

Goodnow

J. J.

(1994). The impact of parental discipline methods on the child’s internalization of values: A reconceptualization of current points of view. Developmental Psychology, 30, 4-19. doi:10.1037/0012-1649.30.1.4

23.

Haddon

(2015). Children’s critical evaluation of parental mediation. Cyberpsychology, 9, Article 2. doi:10.5817/CP2015-1-2

24.

Hartup

W. W.

(1979). The social worlds of childhood. American Psychologist, 34, 944-950.

25.

Hill

N. E.

Bush

K. R.

Roosa

M. W.

(2003). Parenting and family socialization strategies and children’s mental health: Low–Income Mexican–American and Euro–American mothers and children. Child Development, 74, 189-204. doi:10.1111/1467-8624.t01-1-00530

26.

Ito

Horst

Bittanti

Boyd

Herr-Stephenson

Lange

P. G.

. . .Tripp

(2009). Living and learning with new media: Summary of findings from the Digital Youth Project. Cambridge: The MIT Press.

27.

Katz

V. S.

(2010). How children of immigrants use media to connect their families to the community. Journal of Children and Media, 4, 298-315. doi:10.1080/17482798.2010.486136

28.

Kenny

D. A.

Kashy

D. A.

Cook

W. L.

(2006). Dyadic data analysis. New York, NY: Guilford Press.

29.

Kenny

D. A.

Ledermann

(2010). Detecting, measuring, and testing dyadic patterns in the actor-partner interdependence model. Journal of Family Psychology, 24, 359-366. doi:10.1037/a0019651

30.

Kerr

Stattin

Özdemir

(2012). Perceived parenting style and adolescent adjustment: Revisiting directions of effects and the role of parental knowledge. Developmental Psychology, 48, 1540-1553. doi:10.1037/a0027720

31.

Kochanska

(1993). Toward a synthesis of parental socialization and child temperament in early development of conscience. Child Development, 64, 325-347. doi:10.2307/1131254

32.

Kuczynski

(2003). Beyond bidirectionality: Bilateral conceptual frameworks for understanding dynamics in parent-child relations. In Kuczynski

(Ed.), Handbook of dynamics in parent-child relations (pp. 1-24). Thousand Oaks, CA: SAGE.

33.

Kuczynski

De Mol

(2015). Dialectical models of socialization. In Overton

W. F.

Molenaar

P. C. M.

(Eds.), Handbook of child psychology and developmental science: Theory and method (Vol. 1, pp. 326-368). Hoboken, NJ: Wiley.

34.

Kuczynski

Lollis

Koguchi

(2003). Reconstructing common sense: Metaphors of bidirectionality in parent-child relations. In Kuczynski

(Ed.), Handbook of dynamics in parent-child relations (pp. 421-437). Thousand Oaks, CA: SAGE.

35.

Kuczynski

Marshall

Schell

(1997). Value socialization in a bidirectional context. In Grusec

J. E.

Kuczynski

(Eds.), Parenting and children’s internalization of values: A handbook of contemporary theory (pp. 23-50). Hoboken, NJ: John Wiley.

36.

Kuczynski

Pitman

Ta-Young

Harach

(2016). Children’s Influence on parents’ adult development: Mothers’ and fathers’ receptivity to children’s requests for change. Journal of Adult Development, 23, 193-203. doi:10.1007/s10804-016-9235-8

37.

Livingstone

Helsper

E. J.

(2008). Parental mediation of children’s Internet use. Journal of Broadcasting & Electronic Media, 52, 581-599. doi:10.1080/08838150802437396

38.

Livingstone

Mascheroni

Dreier

Chaudron

Lagae

(2015). How parents of young children manage digital devices at home: The role of income, education and parental style. London, England: EU Kids Online, London School of Economics and Political Science.

39.

Livingstone

Ólafsson

Helsper

E. J.

Lupiáñez-Villanueva

Veltri

G. A.

Folkvord

(2017). Maximizing opportunities and minimizing risks for children online: The role of digital skills in emerging strategies of parental mediation. Journal of Communication, 67, 82-105. doi:10.1111/jcom.12277

40.

Maccoby

E. E.

Martin

J. A.

(1983). Socialization in the context of the family: Parent-child interactions. In Mussen

P. H.

(Ed.), Handbook of child psychology, socialization, personality, and social development (Vol. 4, pp. 1-101). New York, NY: Wiley.

41.

Mesch

G. S.

(2006). Family characteristics and intergenerational conflicts over the Internet. Information, Communication & Society, 9, 473-495. doi:10.1080/13691180600858705

42.

Mesch

G. S.

(2009). Parental mediation, online activities, and cyberbullying. Cyberpsychology & Behavior, 12, 387-393. doi:10.1089/cpb.2009.0068

43.

Nathanson

A. I.

(1999). Identifying and explaining the relationship between parental mediation and children’s aggression. Communication Research, 26, 124-143. doi:10.1177/009365099026002002

44.

Nathanson

A. I.

(2001). Parent and child perspectives on the presence and meaning of parental television mediation. Journal of Broadcasting & Electronic Media, 45, 201-220. doi:10.1207/s15506878jobem4502_1

45.

Nathanson

A. I.

(2002). The unintended effects of parental mediation of television on adolescents. Media Psychology, 4, 207-230. doi:10.1207/S1532785XMEP0403_01

46.

Nelissen

Van den Bulck

(2018a). Expanding the unidirectional view on parental television mediation: Children’s guidance of their parent’s television use. Journal of Broadcasting & Electronic Media, 62, 232-250. doi:10.1080/08838151.2018.1451855

47.

Nelissen

Van den Bulck

(2018b). When digital natives instruct digital immigrants: Active guidance of parental media use by children and conflict in the family. Information, Communication & Society, 21, 375-387. doi:10.1080/1369118X.2017.1281993

48.

Nikken

Jansz

(2006). Parental mediation of children’s videogame playing: A comparison of the reports by parents and children. Learning, Media, and Technology, 31, 181-202. doi:10.1080/17439880600756803

49.

Nikken

Jansz

(2014). Developing scales to measure parental mediation of young children’s internet use. Learning, Media, and Technology, 39, 250-266. doi:10.1080/17439884.2013.782038

50.

Nikken

Schols

(2015). How and why parents guide the media use of young children. Journal of Child and Family Studies, 24, 3423-3435. doi:10.1007/s10826-015-0144-4

51.

Olesen

(2000). Childhood, media and viewer positions. In Van den Bergh

Van den Bulck

(Eds.), Children and media: Multidisciplinary approaches (pp. 67-88). Leuven, Belgium: Garant.

52.

Parke

R. D.

(2002). Parenting in the new millennium: Prospects, promises, and pitfalls. In McHale

J. P.

Grolnick

W. S.

(Eds.), Retrospect and prospect in the psychological study of families (pp. 65-93). Mahwah, NJ: Lawrence Erlbaum.

53.

Parsons

(1951). The social system. New York, NY: The Free Press.

54.

Prensky

(2001). Digital natives, digital immigrants. Part 1. On the Horizon, 9, 1-6.doi:10.1108/10748120110424816

55.

Salmon

Nichols

(1983). The next birthday method of respondent selection. Public Opinion Quarterly, 47, 270-276. doi:10.1086/268785

56.

Sameroff

(1975). Transactional models in early social relations. Human Development, 18, 65-70. doi:10.1159/000271476

57.

Sameroff

(2009). The transactional model. In Sameroff

(Ed.), The transactional model of development: How children and contexts shape each other (pp. 247-268). Washington, DC: American Psychological Association.

58.

Saphir

M. N.

Chaffee

S. H.

(2002). Adolescents’ contributions to family communication patterns. Human Communication Research, 28, 86-108. doi:10.1111/j.1468- 2958.2002.tb00799.x

59.

Selwyn

(2004). Exploring the role of children in adults’ adoption and use of computers. Information Technology & People, 17, 53-71. doi:10.1108/09593840410522170

60.

Shin

(2013). Parental socialization of children’s internet use: A qualitative approach. New Media & Society, 17, 649-665. doi:10.1177/1461444813516833

61.

Smetana

J. G.

(1995). Parenting styles and conceptions of parental authority during adolescence. Child Development, 66, 299-316. doi:10.2307/1131579

62.

Stafford

(2003). Parent and sibling interactions during middle childhood. In Vangelisti

A. L.

(Ed.), The Routledge handbook of family communication (pp. 256-270). New York, NY: Routledge.

63.

Valcke

Bonte

De Wever

Rots

(2010). Internet parenting styles and the impact on Internet use of primary school children. Computers & Education, 55, 454-464. doi:10.1016/j.compedu.2010.02.009

64.

Valkenburg

P. M.

Krcmar

Peeters

A. L.

Marseille

N. M.

(1999). Developing a scale to assess three styles of television mediation: “instructive mediation,” “restrictive mediation,” and “social coviewing.” Journal of Broadcasting & Electronic Media, 43, 52-66. doi:10.1080/08838159909364474

65.

Valkenburg

P. M.

Piotrowski

J. T.

Hermanns

de Leeuw

(2013). Developing and validating the Perceived Parental Media Mediation Scale: A self-determination perspective. Human Communication Research, 39, 445-469. doi:10.1111/hcre.12010

66.

Van den Bulck

Custers

Nelissen

. (2016). The child-effect in the new media environment: Challenges and opportunities for communication research. Journal of Children and Media, 10, 30-38. doi:10.1080/17482798.2015.1121897

67.

Van den Bulck

Van den Bergh

. (2000). Parental guidance of children’s media use and conflict in the family. In Van den Bergh

Van den Bulck

(Eds.), Children and media (pp. 131-150). Leuven, Belgium: Garant.

68.

Van den Bulck

Van den Bergh

. (2005). The child-effect in media and communication research: A call to arms and an agenda for research. Communication Yearbook, 29, 35-47. doi:10.1080/23808985.2005.11679043

69.

van Den Eijnden

R. J.

Spijkerman

Vermulst

A. A.

van Rooij

T. J.

Engels

R. C

. (2010). Compulsive Internet use among adolescents: Bidirectional parent-child relationships. Journal of Abnormal Child Psychology, 38, 77-89. doi:10.1007/s10802- 009-9347-8

70.

van Deursen

A. J.

van Dijk

J. A.

Peters

. (2011). Rethinking Internet skills: The contribution of gender, age, education, Internet experience, and hours online to medium-and content- related Internet skills. Poetics, 39, 125-144. doi:10.1016/j.poetic.2011.02.001

71.

Van Rompaey

Roe

Struys

. (2002). Children’s influence on internet access at home: Adoption and use in the family context. Information, Communication & Society, 5, 189-206. doi:10.1108/09593840410522170

72.

Zaman

Nouwen

Vanattenhoven

de Ferrerre

Van Looy

(2016). A qualitative inquiry into the contextualized parental mediation practices of young children’s digital media use at home. Journal of Broadcasting & Electronic Media, 60, 1-22. doi:10.1080/08838151.2015.1127240

73.

Zimmerman

B. J.

(2000). Self-efficacy: An essential motive to learn. Contemporary Educational Psychology, 25, 82-91. doi:10.1006/ceps.1999.1016