Longitudinal Effects of Parent–Child Interactions on Children’s Social Competence

Abstract

Purpose:

This study examined normative change in children’s levels of social competence and parent–child interactions (PCIs) from kindergarten through second grade as well as relations between levels of PCI and children’s social development.

Methods:

Multiple waves of data were collected from parents and teachers of 379 children ranging in age from 4 to 6½ years. Hierarchical linear modeling was used to estimate change over time in social competence controlling for children’s exposure to early learning programs such as home visiting and preschool programs. Next, frequency of PCI was entered as a time-varying predictor of children’s social competence levels.

Results:

Results indicated discrepancies between parent and teacher reports of children’s social development. Parents reported normative growth in children’s social competence, whereas teachers reported declines in this area. Parents also reported decreases in PCIs over time. Parent-reported social competence scores were positively associated with levels of PCIs, whereas no significant association was found between teacher-rated social competence scores and levels of PCI.

Discussion:

These results highlight the importance of viewing children’s social competence as a dynamic variable related to a multitude of individual, familial, and social factors. Findings also emphasize PCIs as a potential target for interventions aimed at enhancing such competence.

Keywords

social competence parent–child interactions school readiness hierarchical linear modeling

Research on school readiness often emphasizes cognitive and academic skills as primary markers of child development. However, a more holistic approach acknowledges social competence as a fundamental aspect of child development and a contributor to school readiness (Morris et al., 2013). Social competence has been conceptualized as a child’s ability to “form close and secure adult and peer relationships; experience, regulate, and express emotions in socially and culturally appropriate ways; and explore the environment and learn” (Yates et al., 2008; p. 2). While the construct remains to be clearly defined in the literature, researchers have focused on social expectations surrounding emotional regulation (e.g., curiosity, attitude toward school, and self-confidence), cooperative interactions with peers and adults (e.g., sharing and taking turns), and self-control (e.g., impulse control and rule following; Eisenberg & Fabes, 2006; Lane, Pierson, & Givner, 2004; Meier, DiPerna, & Oster, 2006). Junttila, Voeten, Kaukiainen, and Vauras (2006) conducted a factor analysis of parent and teacher social competence ratings and identified two primary dimensions of social competence: (1) prosocial behavior encompassing emotional regulation as well as cooperative interactions and (2) antisocial behavior or behavior indicative of poor self-control. Research consistently demonstrates a significant inverse correlation between these two factors (Eisenberg, Smith, & Sprinrad, 2011; Fantuzzo, Manz, & McDermott, 1998; Najaka, Gottfredson, & Wilson, 2001; Vahedi, Farrokhi, & Farrajian, 2012), with some suggesting a direct causal relationship (Langeveld, Gunderson, & Svartdal, 2012). Indeed, Fantuzzo et al. suggested these two subscales represent opposite poles of the overall construct of social competence. Thus, social competence in children is generally considered to involve a number of features but centers around their abilities to meet expectations regarding positive social interactions (i.e., social skills) and the absence of problematic or antisocial behaviors.

Children’s deficits in social competence can present barriers to effective instruction in addition to diverting educational resources away from their primary goal of teaching students. Thus, the Administration for Children and Families includes social competence as a desired outcome for children in Head Start, acknowledging proficiency in this domain to be an integral component of academic achievement (Mattera, Lloyd, Fishman, & Bangser, 2013). Moreover, teachers frequently rank social competence, including minimal levels of problem behavior, above academic skills as contributors to school readiness (Semrud-Clikeman, 2007).

Although research suggests significant normative growth in social competence throughout early and middle childhood (Fine, Izard, & Trentacosta, 2006), children who demonstrate low levels of social skill and high rates of problem behavior in early childhood often continue to exhibit deficits in later years (National Research Council & Institute of Medicine, 2009). Furthermore, children whose parents report low levels of social competence in their interactions at home tend to experience similar difficulties in their interactions at school and with peers (Keane & Calkins, 2004). These difficulties are particularly pronounced among male students and students from disadvantaged backgrounds (Entwisle, Alexander, & Olson, 2005).

Research is needed to identify the specific mechanisms through which growth in social competence might be promoted in early childhood. Specifically, although much research has assessed the contributions of parent–child interactions (PCIs) to children’s social competence in the preschool years (Hart & Risley, 1995; Vahedi et al., 2012), few studies have assessed this relationship for school-aged children. Thus, the current study focuses on PCIs as a predictor of social competence as a facet of school readiness.

Extensive research has focused on children’s early interactions with their primary caregivers as determinants of their emerging social competence (Foster, Lambert, Abbott-Shim, McCarty, & Franze, 2005; McCart, Priester, Davies, & Azen, 2006). In their review of parental participation in Head Start, Henrich and Gadaire (2008) highlighted a number of studies linking parent involvement at home and at school to children’s social competence in each of these settings. For instance, Foster et al. studied the effect of joint activities focused on emergent literacy on preschool children’s social functioning. In a sample of 325 children randomly selected from 48 Head Start classrooms across the country, they identified a significant relationship between children’s participation in home learning activities and their social functioning. This effect was also found to mediate the relationship between socioeconomic status (SES) and school outcomes.

PCIs need not center around educational activities to have a meaningful impact on child development. Using data from the Head Start Family and Child Experiences Survey, O’Brien et al. (2002) asked 120 families participating in Head Start programs across the country how frequently they engaged in various academic and nonacademic activities (e.g., reading a story, completing a puzzle, and attending an athletic event) with their children. Their findings suggested increases in PCIs over the course of a year to be associated with improvements in children’s academic and social competence as well as decreased rates of problem behavior. Similar outcomes were observed by Meek, Robinson, and Jahromi (2011) who found the quality of PCIs in initial assessments to be predictive of social interactions with peers 1 year later. Based on these studies among others, Henrich and Blackman-Jones (2006) proposed a model outlining the primary ways in which parent involvement affects preschool children’s educational outcomes. One pathway in this model involves an indirect effect of PCIs through their effects on children’s social behavior. Increased levels of parental involvement at home and at school were hypothesized to increase children’s social skills and motivation to learn in school. This logic and supporting research has recently led to large-scale public health efforts to increase the frequency with which parents interact with their children (The White House, 2014; Neighmond, 2014).

Cross-sectional analyses have demonstrated that children who enjoy high levels of interactions with their parents exhibit higher levels of social competence than their peers with less involved parents (Foster et al., 2005; Isley, O’Neil, Clatfelter, & Parke, 1999; Valentino, Comas, & Nuttall, 2014). However, inconsistent definitions of PCI and measures of child outcomes across studies make it difficult to compare effect sizes between study samples. Also, longitudinal changes in PCIs may contribute to variable effects on children’s social development over the early school years. When such studies are conducted, they typically treat parenting factors as static states that are maintained invariant throughout the life of the study. A more dynamic approach would incorporate fluctuations in parenting variables (e.g., improvements resulting from intervention programs) as well as the magnitude of these influences given changing family dynamics, child characteristics, and social demands over time (Bronfenbrenner, 1999; Nelson, 2015; Zimmer-Gembeck & Thomas, 2010). In a notable illustration of this approach, El Nokali, Bachman, and Votruba-Drzal (2010) presented a multilevel model of the effects of parent involvement in their children’s educational activities on emerging social competence. These authors found increases in parent involvement from first through fifth grade to be associated with increases in their children’s social competence. Unfortunately, this study did not specifically isolate the effect of PCIs, as it included parent involvement in school activities with as well as without the child present. It is also unclear whether such outcomes were directly related to parents’ involvement in their children’s education or if other PCIs might have similar effects on children’s social development. Research assessing normative changes in PCIs (related to educational as well as leisure activities) in early childhood as well as the effects of these changes on children’s social competence could provide valuable information for family-focused intervention efforts.

The current study sought to identify normative patterns of change in children’s social competence and PCIs in the early school years as well as effects of changes in PCIs on emerging social competence. Research questions were as follows:

Research Question 1: How do children’s levels of social competence change from the time they enter kindergarten through completion of the second grade?

Research Question 2: How do children’s levels of interaction with their parents change over this time period? and

Research Question 3: Are changes in PCIs over time associated with corresponding changes in social competence?

Based on previous research, it was hypothesized that (1) social competence would increase in a linear fashion from kindergarten through second grade, (2) PCIs would decline given children’s increased involvement in school activities and social interactions with peers, and (3) changes in PCI over time would be associated with similar changes (in the same direction) in children’s levels of social competence. These hypotheses were tested in a sample of children enrolled in a national evaluation of the School of the 21st Century (21C), a comprehensive model of education incorporating school-based child care and family support services. Because family support services, such as home visiting, and preschool programs have been found to have effects on both PCIs and children’s social competence (Brooks-Gunn & Markman, 2005; Webster-Stratton, Reid, & Beauchaine, 2011), these experiences were included in initial analyses. Models also took into account demographic factors, such as gender and family SES, that are often associated with both PCIs and children’s social competence (Denham, Bassett, Sirotkin, Brown, & Morris, 2015; Dodge, Coie, & Lynam, 2006; Hill, 2001).

Method

Participants

Children were selected from school districts in five states as part of a national evaluation of the 21C. The 21C is a multisite educational initiative funded through the U.S. Department of Education and various private donors. The program emphasizes overall child development including social and emotional health with a focus on supporting parents and caregivers. Program outcomes were independently evaluated through the Edward Zigler Center on Child Development and Public Policy at Yale University. All procedures were conducted with the approval of the institutional review board (IRB) at Yale University, and IRB approval for secondary analysis of the data was obtained from Georgia State University.

Sites were specifically chosen to represent children living in urban, rural, and suburban settings as well as diverse socioeconomic conditions. Based on the size of the school district, the families of all or a percentage of kindergarten students were recruited to participate in the study. Each school determined its own procedures for distributing and handling returned surveys. To maximize recruitment, schools were provided US$5 for each survey returned.

The parents and teachers of participating students were asked to complete surveys in the fall and spring of the child’s kindergarten year and the spring of the child’s first and second grade years. Thus, the first two waves of data collection were spaced approximately 6 months apart with subsequent waves being separated by 12 months on average. In the fall of the first year, a total of 418 families participated in the kindergarten cohort of the study with 395 (94%) of these families completing parent surveys. In the spring of that year, parent surveys were completed by 310 (74%) of the original participants. Parent surveys were completed by 217 (52%) and 213 (51%) of the original participants in the second and third years of the study, respectively. In the second year of the study, 145 new participants with parent data were added to the sample. Of these participants, 104 (72%) completed parent surveys in the second year and 105 (73%) completed parent surveys in the third and final years of the study.

Of the combined (initial plus refreshed) sample, 117 (21%) of parents completed surveys at all four time periods, 234 (42%) completed surveys at three or more time periods, and 404 (72%) completed surveys at two or more time periods. Analyses were conducted with families who responded at two or more time periods, resulting in a final sample size of 379 who provided data for all level-2 variables. This criterion was established to allow for the inclusion of the greatest number of participants possible without sacrificing the accuracy of random effect predictions.

At the beginning of the study, children ranged in age from 4 years to 6 years and 8 months (M = 5 years, 7 months, SD = 4 months). Fifty-two percent of the students were female and 48% were male. Of the parents who responded to the surveys, 91% identified themselves as the mother of the student, 7% as fathers, and the remainder identified as grandparents, stepparents, or guardians. Participating parents were also asked to report their annual household incomes and educational histories. Specifically, participants were asked to indicate whether their annual household income fell below US$20,000, between US$20,000 and US$39,999, or US$40,000 and above. Participants also reported their highest level of education: high school or less, some college experience, or college graduate. For all analyses, parents with some college experience and parents who had earned college degrees were combined into a single category. The sample was socioeconomically diverse, with 25% of parents reporting annual household incomes less than US$20,000 and 47% reporting annual incomes greater than US$40,000. Sixty-five percent of parents reported having attended college and approximately half of these parents held college degrees. Seventy-eight percent of parents identified themselves as European American, 18% Latino, 3% African American, and 1% of mixed ethnicity.

Students who participated in an early intervention program involving home visitation comprised 48% of the sample. Participation in these programs consisted of visits by a social service professional or nurse associated with any organized early intervention program at any point within a child’s first 3 years. Students who participated in preschool programs comprised 69% of the sample. Sixty-three percent of these children attended preschools based on the 21C model of comprehensive early childhood services. The remaining 37% of children attended other center-based educational programs and/or Head Start prior to beginning kindergarten. Prior research on the 21C evaluation found no significant differences in terms of children’s academic and social functioning or parental involvement in school when comparing 21C preschool to other center-based preschool models, suggesting that the 37% of students who attended preschool elsewhere received similar services to those served in 21C preschools (Henrich, Ginicola, & Finn-Stevenson, 2006). Thus, all preschool experiences were collapsed into a single measure of preschool participation.

Measures

Parents completed surveys in the fall and spring of the first year and in the spring of the next 2 years. Surveys were provided in English and in Spanish with 42 (11%) of the original 393 participants completing the Spanish version. The parent survey included a number of questions that were adapted from several widely used measures of parent involvement (El Nokali, Bachman, & Votruba-Drzal, 2010; West, Denton, & Germino-Hausken, 2000). Survey questions inquired about past participation in home-visiting and preschool programs as well as levels of PCI in the home. Survey items constituted an internally consistent scale of PCI (8 items, Cronbach’s α = .78) including such activities as singing songs, reading books, playing games, and going to community events together. Parents were asked how often they engaged in these behaviors with their children: never or hardly ever, once or twice a month, once or twice per week, or almost every day. At Time 1, parents reported interaction scores ranging from 8 to 32, the maximum score possible. The mean PCI score was 25.43 (SD = 3.71).

Parents were asked to report on their children’s social and behavioral functioning using questions from the Social Skills Rating System (SSRS; Gresham & Elliott, 1990). The Social Skills domain of the SSRS consists of 16 questions designed to assess four dimensions of prosocial behavior: Cooperation, Assertion, Responsibility, and Self-control. Questions from this domain ask parents to rate how often (never, sometimes, and very often) their child demonstrates specific skills such as appropriately expressing his or her feelings, joining group activities without being told to, and ending disagreements calmly. The Problem Behaviors domain includes 19 questions divided into three subscales: Hyperactivity, Internalizing Symptoms, and Externalizing Symptoms. On the same 3-point scale, parents are asked to indicate how often their child engages in such problematic behaviors as disobeying rules or requests, having temper tantrums, and acting sad or depressed. Reliability coefficients (Cronbach’s α) for the Social Skills and Problem Behaviors domains were .77 and .75, respectively, at Time 1.

The SSRS has been found to have satisfactory reliability and validity (Gresham & Elliott, 1990). However, low convergent validity has been observed between the parent and teacher forms. Such lack of agreement might be expected given the variation in children’s social behavior across contexts and disparities in the content of the two versions. Thus, to promote the generalizability of findings to school as well as home environments, the teachers of participating students were recruited to complete the SSRS at each wave of data collection. The format of the teacher form is identical to that of the parent form although some questions are adapted to assess behaviors specific to the classroom setting. For instance, teachers are asked to rate how often students volunteer to help peers with classroom tasks, accept peers’ ideas for group activities, and use free time in acceptable ways. Like the parent form of the survey, the teacher form was found to assess internally consistent dimensions of Social Skills (30 questions, α = .92) and Problem Behaviors (18 questions, α = .91). In the first year of data collection, 34 teachers completed the SSRS for participating children and response rates ranged from 93% to 97% over the course of the study. Thus, it is estimated that teacher ratings were obtained for over 90% of participating children at each time point.

Overall social competence scores were calculated by adding children’s composite scores in the Social Skills domain with their reverse coded composite scores in the Problem Behavior domain. Although formulation of this construct is largely inconsistent in the extant literature, clear support exists for including both prosocial and problem behaviors in a comprehensive measure of social competence. With regard to the SSRS in particular, studies validating its psychometric properties suggest that the Prosocial and Problem Behavior subscales represent opposite ends of the same general construct (Fantuzzo et al., 1998). Using hierarchical and canonical analyses, these authors found the two subscales did not relate differentially to external psychometric criteria. Similarly, Vahedi, Farrokhi, and Farajian (2012) reported concomitant increases in both Prosocial and Problem Behavior scores associated with intervention (Head Start). Thus, it was considered most parsimonious to incorporate both scores into a single dependent variable.

Analyses

Analyses utilized hierarchical linear modeling (HLM 5; Raudenbush & Bryk, 2002) using full information maximum likelihood estimation. First, patterns of missing data were probed. Because the sample was refreshed in the second year of the study, the number of time periods in which data were collected varied across participants. In other words, a fair amount of missing data was inherent in the study design. HLM accommodates this imbalance by estimating growth parameters based on the data available for each individual as well as a model-based trajectory estimated from values observed in the larger sample. However, to ensure the validity of generalization from observed data, patterns of missing data were explored to ensure the absence of systematic, theoretically significant bias. Specifically, linear regression analyses were conducted to predict the number of time points for which each participant was missing data. These analyses indicated that the number of waves for which a participant was missing data could be predicted based on observed data values of variables included in the initial models. Specifically, the amount of missing data was predictable based on children’s preschool participation (β = .12, SE = .04, p < .01), parental education levels (β = −.26, SE = .10, p < .01), and household income (β = −.24, SE = .08, p < .01).

In longitudinal analyses, multilevel modeling permits the estimation of both within-subject and between-subject variation in specified outcomes. Analysis begins with the estimation of unconditional growth models to predict the average level of change expected across the entire sample of interest (without accounting for covariates or effects of independent variables). Next, a series of conditional growth models are estimated to assess the effects of these variables. The first level of analysis represents within-subject change over time and includes repeated measures of variables that are expected to fluctuate over the course of the study. The second level of analysis represents between-subject differences that are associated with fixed predictors, that is, predictors expected to remain constant over time. Multilevel modeling facilitates assessment of the effects of these predictors on variability in both the Level-1 intercept and slope. Specific control variables entered at Level 2 included student gender, parental education level, household income, participation in home visiting programs, and preschool participation, and all reported standard errors refer to robust standard errors.

To address Research Question 1, Level 1 contained repeated measures of children’s parent- and teacher-rated social competence scores, which were estimated in separate analyses (see Equation 1). The intercept in this model was centered to represent the average level of social competence reported at Time 2 (73 months; spring of child’s kindergarten year). This time period was selected because it was the period at which data were collected for the largest number of children. The slope is estimated as the degree of change per month (an estimate that remains constant from kindergarten through second grade). Variability in initial levels of social competence as well as rates of change over time was modeled at Level 2 using Equations 2 and 3, respectively. Thus, these equations allowed for more precise estimation of change over time in social competence accounting for the effects of relevant covariates.

Social competence = π_{0} {+ π}_{1} Age + e,

\begin{array}{l} π_{0} = γ_{00} + γ_{01} Gender + γ_{02} Education + γ_{03} Income \\ + γ_{04} Homevisits + γ_{05} Preschool + r, \end{array}

\begin{array}{l} π_{1} = γ_{10} + γ_{11} Gender + γ_{12} Education + γ_{13} Income \\ + γ_{14} Homevisits + γ_{15} PreSchool + r . \end{array}

To address Research Question 2, PCIs were estimated using similar models. In Equation 4, the intercept represents average rates of PCIs at Time 2 (73 months; spring of the child’s kindergarten year), and the slope represents the rate of change in PCIs per month. Variability in initial levels and rates of change over time were modeled at Level 2 using Equations 5 and 6, respectively.

PCIs = π_{0} {+ π}_{1} Age + e,

\begin{array}{l} π_{0} = γ_{00} + γ_{01} Gender + γ_{02} Education + γ_{03} Income \\ + γ_{04} Homevisits + γ_{05} Preschool + r, \end{array}

\begin{array}{l} π_{1} = γ_{10} + γ_{11} Gender + γ_{12} Education + γ_{13} Income \\ + γ_{14} Homevisits + γ_{15} PreSchool + r . \end{array}

Research Question 3 sought to identify the effects of changes in PCIs on children’s social competence over time. Thus, PCIs were added as a time-varying predictor to the Level-1 equation predicting children’s levels of social competence (see Equation 7). For this analysis, Level-2 equations were required to estimate: the Level-1 intercept (average level of social competence at Time 2), the coefficient associated with Age (rate of change in social competence per month), and the coefficient associated with PCIs (rate of change in social competence per unit increase in PCIs). Level-2 equations associated with this analysis are depicted in Equations 8 –10.

{Social competence = π}_{0} {+ π}_{1} {Age + π}_{2} PCIs + e,

\begin{array}{l} π_{0} = γ_{00} + γ_{01} Gender + γ_{02} Education + γ_{03} Income \\ + γ_{04} Homevisits + γ_{05} Preschool + r, \end{array}

\begin{array}{l} π_{1} = γ_{10} + γ_{11} Gender + γ_{12} Education + γ_{13} Income \\ + γ_{14} Homevisits + γ_{15} PreSchool + r . \end{array}

\begin{array}{l} π_{2} = γ_{20} + γ_{21} Gender + γ_{22} Education + γ_{23} Income \\ + γ_{24} Homevisits + γ_{25} PreSchool + r . \end{array}

Covariates with statistically insignificant effects were systematically removed from the above models using change in deviance tests. All deviance comparisons were made using the same number of respondents. When a variable was removed with no significant change in model deviance, the simpler model was considered to be more parsimonious without sacrificing overall model fit.

Results

At Time 1, children’s parent-reported scores in the Social Skills domain ranged from 2 to 36 of a possible 48 points (M = 26.2, SD = 5.29). Children’s scores in the Problem Behaviors domain ranged from 0 to 28 of a possible 38 points (M = 8.8, SD = 4.46). Overall social competence scores ranged from −10 to 32 (with −38 being the lowest possible score and 48 being the highest possible). The average child was reported to have a social competence score of 15.31 (SD = 8.52). Teacher-rated social skills scores ranged from 7 to 58 (M = 41.22, SD = 9.74) of a total of 60 possible points. Problem behavior scores ranged from 0 to 31 of a possible 36 points (M = 6.69, SD = 6.85). Given a possible range from −36 to 60, overall teacher-rated social competence scores ranged from −22 to 56. The average teacher-rated social competence score was 34.52 (SD = 15.57).

Table 1 presents the correlations among variables included in each of the models. Correlations were derived from student information provided in the spring of their kindergarten year (N = 379) to provide a cursory impression of the associations between variables. First, a moderate direct correlation was found between parent-reported social competence and teacher-reported social competence. Parent-reported social competence was also significantly positively correlated with PCIs, parent education level, and household income. Of note, parent-reported social competence scores did not correlate significantly with children’s participation in either home visiting or preschool programs.

Table 1.

Correlations Among Variables in the Spring of Children’s Kindergarten Year.

Variable	Social Competence (Parent)	Social Competence (Teacher)	PCI	Preschool	Home Visits	Income	Parent Education	Child Sex
Parent reported social competence	1.0	0.33**	0.18*	.03	.01	.23**	.17**	−.11
Teacher reported social competence	0.33**	1.0	0.08	.06	−.12**	.09*	−.01	−.24**
PCI	0.18**	0.08	1.0	.01	.08	.18**	.19**	.01

Note. PCI = parent–child interaction.

*p < .05. **p < .01.

Teacher-reported social competence was significantly correlated with household income and child sex, with female students receiving higher ratings from their teachers than male students. Past participation in home visiting programs was negatively correlated with teacher-reported social competence, suggesting that children who participated in such programs tended to receive lower social competence ratings from their teachers. No statistically significant association was found between preschool participation and teacher-reported social competence.

Levels of PCI were positively correlated with parent education level and household income, though no significant relation was found between PCIs and participation in either home visiting or preschool programs.

Development of Social Competence Over Time

To address Research Question 1, unconditional growth models were created to estimate trajectories of change in parent- and teacher-reported social competence for each student, as they progressed from kindergarten through third grade (N = 379). These models included only random effects at Level 2. At 73 months of age, the average child obtained a parent-reported social competence score of 16.94 (SE = .41, p < .01), which tended to increase over time (γ₁₀ = .18, SE = .01, p < .01). There was significant residual variance at Level 2 for both intercept (σ₀ ² = 54.61, p < .01) and slope (σ₁ ² = 0.02, p < .01). Teacher reports revealed a slightly different picture. Based on teacher reports, the mean social competence score at 73 months of age was 35.26 (SE = .69, p < .01) which tended to decrease over time (γ₁₀ = −.11, SE = .03, p < .01). There was significant residual variance at Level 2 for intercept (σ₀ ² = 127.64, p < .01) and slope (σ₁ ² = 0.09, p < .01).

Level-2 fixed effects

Covariates (child sex, parental education level, household income, preschool participation, and participation in home visiting programs) were then added to the unconditional growth models described above (see Table 2). There were significant effects of child sex, parent education, and household income on the intercept of parent-reported social competence. The introduction of these variables to the model resulted in a significant reduction in model deviance, χ² ₍₄₎ = 46.67, p < .01, indicating that their inclusion improved model fit. Female students were rated as more socially competent than male students in the spring of their kindergarten year. Students whose parents reported some amount of postsecondary education were also rated as having higher initial levels of social competence as were students whose parents reported higher than average household incomes.

Table 2.

Main Effects of Control Variables.

Variable	Parameter	Parent-Rated Social Competence (SE)	Teacher-Rated Social Competence (SE)	Parent–Child Interactions (SE)
N		379	379	379
Fixed effects
For intercept
Intercept	γ₀₀	15.39** (−1.18)	40.69** (0.96)	25.43** (0.15)
Child sex	γ₀₁	−1.63* (0.79)	−7.82** (1.31)	—
Parent education	γ₀₂	1.17* (0.55)	—	—
Income	γ₀₃	2.16** (0.58)	2.55** (0.78)	0.60** (0.21)
Home visits	γ₀₄	—	−3.44** (1.27)	—
For age
Intercept	γ₁₀	0.18** (0.01)	−0.16* (0.07)	−0.05** (0.01)
Child sex	γ₁₁	—	−0.15* (0.06)	—
Income	γ₁₃	0.04* (0.02)	—	−0.03 (0.01)
Preschool	γ₁₅	—	0.06* (0.03)	—
Variance components
Level 1: Within person	σ_∊ ²	20.10 (4.48)	118.28 (10.88)	4.07 (2.02)
Level 2: Intercept	σ₀ ²	49.61** (7.04)	103.79** (10.19)	7.02** (2.65)
Age	σ₁ ²	0.02** (0.12)	0.08** (0.29)	0.01** (0.08)

*p < .05. **p < .01.

Rates of change did not vary significantly by parental education or child sex, and removal of these variables from the model resulted in minimal change in model deviance, χ² ₍₂₎ = 1.02, p > .05, indicating that removing these effects led to a more parsimonious model without an impact on model fit. Significant residual variance indicated that added predictors at Level 2 could improve the model’s ability to accurately predict initial levels (σ₀ ² = 49.61, p < .01) and rates of change (σ₁ ² = 0.02, p < .01) in children’s parent-reported social competence scores. Greater rates of growth were reported for students whose parents reported higher incomes. Significant residual variance indicated that added predictors at Level 2 could improve the model’s ability to accurately predict initial levels and rates of change in children’s parent-reported social competence scores.

Regarding teacher-reported social competence, significant effects were observed for child sex and household income in the model for intercept and child sex in the model for slope. The introduction of these variables resulted in a notable reduction in model deviance χ² ₍₃₎ = 63.02, p < .01. Specifically, teachers-rated female students significantly higher than male students in initial social competence, and they reported larger declines in boys’ social competence scores relative to girls’ social competence scores. Further, children from families with above average incomes were rated as having greater social competence at 73 months than children from families with average and below average household incomes.

Inclusion of variables representing children’s participation in home visiting programs (intercept model) and preschool (slope model) also led to a significant improvement in model fit, χ² ₍₂₎ = 12.09, p < .01. Children who participated in home visiting programs tended to enter school with lower social competence ratings than children with no home visiting experience. Teachers reported that children who attended preschool exhibited lower rates of decline in social competence than those who had not attended preschool.

Variables representing parent education and children’s preschool participation (intercept model) as well as parent education, household income, and home visits (slope model) did not have a significant impact on model fit, χ² ₍₅₎ = 1.51, p > .05, and were thus removed from the model for future analyses. The residual variance at Level 2 remained significant for intercept and slope suggesting the possibility of predicting increased between-subject variance through the inclusion of additional predictors at Level 2.

In summary, children’s parent-reported social competence increased over time as hypothesized. In contrast to parent-reported social competence, teacher-reported social competence tended to decline over this same time period.

PCIs

To address the second research question, an unconditional growth model (including only random effects at Level 2) was used to predict PCI levels. Based on this model, average levels of interaction were found to be high at 73 months of age (γ₀₀ = 25.43, SE = .16, p < .01) but tended to decline significantly and steadily between kindergarten and second grade (γ₁₀ = −.05, SE = .01, p < .01). Additionally, there was significant residual variance at Level 2 (σ₀ ² =7.29, p < .01; σ₀₁ ² = 0.01, p < .01).

Level-2 fixed effects

The inclusion of household income in models for both intercept and slope resulted in a reduction in model deviance, χ² ₍₂₎ = 15.22, p < .01. No significant differences were found relative to child sex, parental education, participation in preschool, or home-visiting programs, and these variables did not contribute significantly to model fit, χ² ₍₈₎ = 6.76, p > .05. Higher rates of interaction were observed in families reporting higher than average household incomes. The rate of decline in PCIs tended to be lower in families earning above average incomes. The possibility of improving model fit through added Level 2 predictors was indicated by significant residual variance at Level 2 for both intercept and slope. In summary, levels of PCIs tended to be high upon school entry but decreased over time as hypothesized.

Effects of PCIs on Social Competence

The third question of interest in this study pertained to the relationship between PCIs and children’s social competence. Given that PCI levels could change from one data collection period to the next, this variable was considered a time-varying predictor and was entered at Level 1. In these models, Level 2 included the same variables as in previous analyses. Results are reported in Table 3.

Table 3.

Effects of PCI on Children’s Social Competence.

Variable	Parameter	Parent-Rated Social Competence (SE)	Teacher-Rated Social Competence (SE)
n		379	367
Fixed effects
For intercept
Intercept	γ₀₀	16.25** (−0.76)	41.56** (0.55)
Child sex	γ₀₁	−1.63* (0.79)	−8.42** (1.24)
Parent education	γ₀₂	2.15* (0.92)	—
Income	γ₀₃	2.26** (0.54)	1.92* (0.78)
Home visits	γ₀₄	—	−3.99* (1.26)
For age
Intercept	γ₁₀	0.19** (0.02)	−0.09** (0.03)
For PCI
Intercept	γ₂₀	0.11* (0.05)	0.33 (0.34)
Variance components
Level 1: Within person	σ_∊ ²	19.5 (4.42)	113.98 (10.68)
Level 2: Intercept	σ₀ ²	48.25** (6.95)	97.23** (9.86)
Age	σ₁ ²	0.02** (0.14)	0.06** (0.24)

Note. PCI = parent–child interaction.

*p < .05. **p < .01.

Based on parent ratings, increased PCIs were significantly and positively associated with children’s social competence scores over time. The addition of this variable resulted in a significant increase in model fit, χ² ₍₁₎ = 138.85, p < .01. The absence of a significant random effect for PCIs indicated the relationship between changes in PCIs and children’s social competence scores tended to be uniform across participants. When this random effect was removed, no significant change was observed in the model deviance indicating the inclusion of this term did not add to the model’s goodness of fit, χ² ₍₃₎ = 7.20, p > .05. Additionally, none of the Level-2 fixed effects (parent education level, household income, child sex, or children’s participation in home visiting or preschool programs) on the slope of PCIs was significant, indicating that it was not moderated by the Level-2 variables. The removal of these variables from the model predicting the slope of PCIs did not significantly affect the model’s goodness of fit, χ² ₍₅₎ = 5.34, p > .05. However, the residual variance of age at Level 2 remained significant, suggesting the possibility of predicting increased between-subject variance in the rate of growth in social development through the inclusion of additional predictors at Level 2.

According to teachers, PCIs were not significantly associated with children’s levels of social competence. There was not a significant random effect for the slope of PCIs, indicating that the lack of an effect on levels of teacher-rated social competence over time was uniform across the sample. Indeed, there were no fixed effects of any of the Level-2 variables on the slope of PCIs. Removal of these variables resulted in a minimal change in model fit, χ² ₍₅₎ = 3.98, p > .05. The addition of PCIs to the model at Level 1 resulted in the effects of child sex and preschool participation no longer being significant in the slope model. Thus, these variables were removed from the model with minimal change in model fit, χ² ₍₂₎ = 4.61, p > .05. As in the model predicting parent-rated social competence, statistically significant residual variance at Level 2 indicates the possibility of predicting increased between-subject variance in the rate of social development through the inclusion of additional predictors at Level 2 for both intercept and age.

In sum, parent and teacher reports were inconsistent regarding the relationships between PCIs and social competence. Specifically, higher PCIs were associated with more parent-reported social competence over time but not with teacher-reported social competence scores.

Interaction Between PCIs and Age

Finally, an exploratory analysis was conducted to assess whether the effects of PCIs would differentially predict children’s rate of growth in social competence over time. This model was created by adding an interaction term at Level 1 to represent the product of PCI scores and children’s ages. Results suggested that the effects of PCIs on social competence were constant over time. When this interaction term was removed from the model predicting parent-rated social competence scores, no significant change was noted in model deviance, χ² ₍₁₎ = 0.03, p > .05. Removal of the same interaction term from the model predicting teacher-reported social competence scores resulted in a similarly insignificant change in model deviance, χ² ₍₁₎ = 0.01, p > .05. Lack of an interaction also indicated that the positive effect of PCIs on social competence was constant over time.

Discussion and Applications to Practice

The goal of the present study was to assess the development of social competence over the early school years with emphasis on the association between PCIs and social development. The finding that parent-reported social competence increased over time supports previous research indicating normative growth in this domain (Fine et al., 2006). It is also not surprising that parents and teachers rated children as demonstrating varying levels of social competence since child behavior has been shown to vary between home and school settings (De Los Reyes, Thomas, Goodman, & Kundey, 2013; Wakschlag et al., 2008). Although the finding that parent- and teacher-reported social trajectories followed opposing directions was not anticipated, similar results were found in at least one previous study in which the SSRS was used to assess change in social skill (but not problem behavior) over time (Chan, Ramey, Ramey, & Schmitt, 2000). It is possible that both studies highlight children’s difficulty keeping pace with the rapidly increasing social demands and expectations of students in school over these years. As children progress through school, certain skills may become increasingly relevant in that context but not necessarily in the context of family interactions. For example, qualities that are viewed favorably in the home environment (e.g., autonomy and assertiveness) may not be as desirable in classroom settings, whereas qualities such as cooperation and compliance may become increasingly important. Increasing academic demands in school may also (1) evoke problem behaviors that are not an issue in the less-structured home environment or (2) reduce teachers’ opportunities to observe emerging social competencies. Thus, teachers may have varying expectations of students in kindergarten, first and second grades in addition to observing behavior changes in response to increasing demands, whereas parents may be more sensitive to the development of social competence in their children.

The observed decrease in PCIs over the early school years is consistent with prior research showing that parent involvement in school also tends to decrease over time (Henrich & Blackman-Jones, 2006). It is possible that children’s increasing autonomy and a greater emphasis on peer relationships are related to decreases in parent involvement both at school and at home. Further, children’s school attendance may reduce opportunities for PCI given that they are away from home for more hours each day. However, these hypotheses have not yet been empirically studied, and further research is needed to assess the reasons for observed changes in parents’ behavior as well as potential effects of such changes on child outcomes. Whatever the reason, these findings underscore the importance of viewing family influences as dynamic rather than static predictors of child outcomes.

Results indicating a positive association between levels of PCIs and children’s parent-reported social competence are consistent with earlier studies (Foster, Lambert, Abbott-Shim, McCarty, & Franze, 2005; Meek, Robinson, & Jahromi, 2011). Nye, Turner, and Schwartz (2006) provide a synthesis of several longitudinal studies that have shown relationships between academic achievement and participation in programs designed to increase the quality and/or frequency of PCIs. Similarly, improvements in the quality of PCIs have been associated with decreases in children’s problem behavior (Thomas & Zimmer-Gembeck, 2007). However, few of these studies directly assess changes in the frequency of PCIs, and none focused on changes in children’s overall social competence. By employing HLM, the current investigation serves as a systematic replication of El Nokali et al. (2010), illustrating increases in a variety of PCIs to be associated with gains in children’s social competence. As such, the current findings reinforce the need for early childhood programs involving effective family-focused interventions for enhancing children’s social competence.

Other interesting findings that emerged involved the relation between target outcomes and demographic variables such as child sex, household income, and parental education as well as participation in early childhood programs. First, results indicating higher social competence scores for children whose parents attended college or earn above average incomes highlight the relevance of social and economic policies addressing these factors to children’s lives. The finding that children whose families earn lower incomes tend to develop social competence at a slower rate than their more economically advantaged peers demonstrates the ongoing influence of financial inequality on children’s development. Also, while results indicating sex-based differences in social competence are consistent with past research (Gresham & Elliott, 1990; Keane & Calkins, 2004), to our knowledge this study is unique in identifying substantive declines in social competence among male students. These findings suggest male students could benefit from increased emphasis on social competence in home and school settings to prevent such deterioration. Finally, higher levels of PCIs were reported among families with higher household incomes. It may be that parents with higher earning careers also have more time to engage in activities with their children whereas parents who are struggling financially are likely to be working as much as possible to support their families. Regarding children’s participation in home-visiting and preschool programs, it is important to note that these variables included a broad range of experiences with no objective measure of the regularity of participation or the quality of services received. Children who participated in programs of poor quality for only brief periods of time would likely demonstrate minimal change in social competence or PCIs (Krzewina, 2012; Melhuish, 2001). Further, results of teacher reports indicating slightly lower levels of initial social competence among students who participated in home-visiting programs suggest the possibility that these differences predated students’ receipt of home visits and may have been contributing factors in their referral for these services. Taken together, findings related to family demographic characteristics and participation in early childhood programs support viewing children’s social development through an ecological framework consisting of myriad direct and indirect influences (Bronfenbrenner, 1979).

In sum, these findings are consistent with an ecological framework in which children’s social development is viewed as a dynamic process related to a multitude of individual, familial, and social factors. Moreover, since environmental influences such as PCIs are themselves subject to change systematically over time, further research is needed to better understand how these changes relate to differential outcomes for children. It is particularly noteworthy that, although PCIs tended to decrease over time, findings also indicated that when parents increased their PCIs, children’s social competence also tended to increase. Based on these findings, it is likely children would benefit from early education programs addressing parent involvement at home as well as the broader socioeconomic conditions in which children are raised. Results indicating declines in children’s social competence in school as well as PCIs at home also underscore the importance of early intervention in each of these areas.

Several limitations of the current study suggest avenues for continued investigation. Specifically, children from diverse ethnic backgrounds were underrepresented in the current sample limiting the generalizability of the study’s findings to these populations. Future research should assess the ways in which ethnicity is likely to influence definitions of social competence, children’s development of social competence, as well as relationships between parenting and social development. Additionally, even though household income was included as a control variable in all analyses, the limited range of income observed substantially decreased the likelihood of finding a significant association between income and child outcomes. Thus, the moderating role of family income deserves further exploration. Finally, the current study focused on the effects of PCI frequency on children’s social competence, whereas the type and quality of such interactions are likely to play a significant role in children’s social development. What’s more, the models employed did not test for nonlinear change or reciprocal effects of PCIs and social competence over time. Thus, we cannot rule out the possibility that PCIs increased as a function of children’s emerging social competence. Indeed, children’s skill development may serve to reinforce their parents’ efforts at social engagement and provide greater opportunities for stimulating interaction.

Further research is needed to identify characteristics of PCIs that tend to be associated with increases in children’s social competence over time and normative changes in these characteristics over the early childhood years. Family intervention programs would benefit greatly from increased knowledge of the interactive effects of changes in both quality and quantity of PCIs over time. Such knowledge can assist clinicians in identifying the necessary and sufficient components of intervention programs to maximize positive outcomes for children.

Footnotes

Authors’ Note

This article is based on a thesis submitted by the first author under the supervision of the second author in partial fulfillment of the requirements for the master’s degree in psychology. We thank committee members Gabriel Kuperminc and Lisa Armistead for their helpful comments on previous versions of this article.

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 authors received no financial support for the research, authorship, and/or publication of this article.

References

Bronfenbrenner

(1979). The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press.

Bronfenbrenner

(1999). Environments in developmental perspective: Theoretical and operational models. In Friedman

S. L.

Wachs

T. D.

(Eds.), Measuring environment across the life span: Emerging methods and concepts (pp. 3–28). Washington, DC: American Psychological Association.

Brooks-Gunn

Markman

(2005). The contribution of parenting to ethnic and racial gaps in school readiness. The Future of Children, 15, 139–168. doi:10.1353/foc.2005.0001

Chan

Ramey

Schmitt

(2000). Modeling intraindividual changes in children’s social skills at home and at school: A multivariate latent growth approach to understanding between-settings differences in children’s social skills development. Multivariate Behavioral Research, 35, 365–396. doi:10.1207/S15327906MBR3503_04

De Los Reyes

Thomas

S. A.

Goodman

K. L.

Kundey

S. M. A.

(2013). Principles underlying the use of multiple informants’ reports. Annual Review of Clinical Psychology, 9, 123–149. doi:10.1146/annurev-clinpsy-050212-185617

Denham

S. A.

Bassett

H. H.

Sirotkin

Y. S.

Brown

Morris

C. S.

(2015). “No-o-o-o Peeking”: Preschoolers’ executive control, social competence, and classroom adjustment. Journal of Research in Childhood Education, 29, 212–225. doi:10.1080/02568543.2015.1008659

Dodge

K. A.

Coie

J. D.

Lyman

(2006). Aggression and antisocial behavior in youth. In Damon

Lerner

R. M.

(Series Eds.) Eisenberg

(Vol. Ed.), Handbook of child psychology: Vol. 3. Social, emotional, and personality development (6th ed., pp. 719–788). New York, NY: Wiley.

Eisenberg

Fabes

R. A.

(2006). Emotion regulation and children’s socio-emotional competence. In Balter

Tamis-LeMonda

(Eds.), Child psychology: A handbook of contemporary issues (2nd ed., pp. 357–381). New York, NY: Psychology Press.

Eisenberg

Smith

C. L.

Spinrad

T. L.

(2011). Effortful control: Relations with emotion regulation, adjustment, and socialization in childhood. In Vohs Baumeister (Eds.), Handbook of self-regulation: Research, theory, and applications (2nd ed., pp. 263–283). New York, NY: Guilford Press.

10.

El Nokali

N. E.

Bachman

H. J.

Votruba-Drzal

(2010). Parent involvement and children’s academic and social development in elementary school. Child Development, 81, 988–1005. doi:10.1111/j.1467-8624.2010.01447.x

11.

Entwisle

D. R.

Alexander

K. L.

Olson

L. S.

(2005). First grade and educational attainment by age 22: A new story. American Journal of Sociology, 110, 1458–1502. doi:10.1086/428444

12.

Fantuzzo

Manz

P. H.

McDermott

(1998). Preschool version of the social skills rating system: An empirical analysis of its use with low-income children. Journal of School Psychology, 36, 199–214. doi:10.1016/S0022-4405(98)00005-3

13.

Fine

Izard

Trentacosta

(2006). Emotion situation knowledge in elementary school: Models of longitudinal growth and preschool correlates. Social Development, 15, 730–751. doi:10.1111/j.1467-9507.2006.00367.x

14.

Foster

M. A.

Lambert

Abbott-Shim

McCarty

Franze

(2005). A model of home learning environment and social risk factors in relation to literacy and social outcomes. Early Childhood Research Quarterly, 20, 13–36. doi:10.1016/j.ecresq.2005.01.006

15.

Gresham

Elliot

(1990). Social skills rating system. Circle Pines, MN: American Guidance Services.

16.

Hart

Risley

T. R.

(1995). Meaningful differences in the everyday experiences of young American children. Baltimore, MD: Brookes.

17.

Henrich

C. C.

Blackman-Jones

(2006). Parent involvement in preschool. In Zigler

Gilliam

W. S.

Jones

S. M.

(Eds.), A vision for universal preschool education (pp. 149–168). Cambridge, MA: Cambridge University Press.

18.

Henrich

C. C.

Gadaire

(2008). Head start and parent involvement. Infants & Young Children, 21, 56–69. doi:10.1097/01.IYC.0000306373.48038.e6

19.

Henrich

C. C.

Ginicola

M. M.

Finn-Stevenson

(2006). The school of the 21st Century is making a difference: Findings from two research studies. New Haven, CT: Yale University.

20.

Hill

N. E.

(2001). Parenting and academic socialization as they relate to school readiness: The roles of ethnicity and family income. Journal of Educational Psychology, 93, 686–697.

21.

Isley

O’Neil

Clatfelter

Parke

(1999). Parent and child expressed affect and children’s social competence: Modeling direct and indirect pathways. Developmental Psychology, 35, 547–560.

22.

Junttila

Voeten

Kaukiainen

Vauras

(2006). Multisource assessment of children’s social competence. Educational and Psychological Measurement, 66, 874–895. doi:10.1177/0013164405285546

23.

Keane

S. P.

Calkins

S. D.

(2004). Predicting kindergarten peer social status from toddler and preschool problem behavior. Journal of Abnormal Child Psychology, 32, 409–423.

24.

Krzewina

J. M.

(2012). Effect of preschool classroom quality on social and language development (Theses and Dissertations Paper 60). Retrieved June 1, 2015, from http://dc.uwm.edu/etd/60

25.

Lane

K. L.

Pierson

M. R.

Givner

C. C.

(2004). Secondary teachers’ views on social competence: Skills essential for success. Journal of Special Education, 38, 174–186.

26.

Langeveld

J. H.

Gunderson

K. K.

Svartdal

(2012). Social competence as a mediating factor in reduction of behavior problems. Scandinavian Journal of Educational Research, 56, 381–399. doi:10.1080/00313831.2011.594614

27.

Mattera

Lloyd

C. M.

Fishman

Bangser

(2013). A first look at the Head Start CARES demonstration: Large scale implementation of programs to improve children’s social-emotional competence (OPRE Report 2013-47). Washington, DC: Office of Planning, Research and Evaluation, Administration for Children and Families, U.S. Department of Health and Human Services. Retrieved April 25, 2015, from http://www.acf.hhs.gov/programs/opre

28.

McCart

M. R.

Priester

P. E.

Davies

W. H.

Azen

(2006). Differential effectiveness of behavioral parent-training and cognitive-behavioral therapy for antisocial youth. A meta-analysis. Journal of Abnormal Child Psychology, 34, 525–541. doi:10.1007/s10802-006-9031-1

29.

Meek

S. E.

Robinson

L. T.

Jahromi

L. B.

(2011). Parent-child predictors of social competence with peers in children with and without autism. Research in Autism Spectrum Disorders, 6, 815–823. doi:10.1016/j.rasd.2011.11.001

30.

Meier

C. R.

DiPerna

J. C.

Oster

M. M.

(2006). Importance of social skills in the elementary grades. Education and Treatment of Children, 29, 409–419. Retrieved from https://www-jstor-org.web.bisu.edu.cn/stable/42899893

31.

Melhuish

E. C.

(2001). The quest for quality in early day care and preschool experiences continues. International Journal of Behavioral Development, 25, 1–6.

32.

Morris

Lloyd

C. M.

Millenky

Leacock

Raver

C. C.

Bangser

(2013). Using classroom management to improve preschoolers’ social and emotional skills: Final impact and implementation findings from the Foundations of Learning Demonstration in Newark and Chicago. New York, NY: MDRC.

33.

Najaka

S. S.

Gottfredson

D. C.

Wilson

D. B.

(2001). A meta-analytic inquiry into the relationship between selected risk factors and problem behavior. Prevention Science, 2, 257–271. doi:10.1023/A:1013610115351

34.

National Research Council & Institute of Medicine. (2009). Preventing mental, emotional, and behavioral disorders among young people. Committee on the Prevention of Mental Health Disorders and Substance Abuse among Children, Youth, and Young Adults: Research Advances and Promising Interventions. In O’Connell

M. E.

Boat

Warner

K. E.

(Eds.), Board on children, youth, and families, division of behavioral and social sciences and education (pp. 16–32). Washington, DC: The National Academies Press.

35.

Neighmond

(2014). For the children’s sake: Put down that smartphone. Shots-Health News from NPR. Retrieved from http://www.npr.org/sections/health-shots/2014/04/21/304196338/for-the-childrens-sake-put-down-that-smartphone

36.

Nelson

J. A.

(2015). Child reactivity moderates the over-time association between mother-child quality and externalizing problems. International Journal of Behavioral Development, 39, 376–382. doi:10.1177/0165025415573643

37.

Nye

Schwartz

Turner

(2006). Approaches to parent involvement for improving the academic performance of elementary school age children. Philadelphia, PA: The Campbell Collaboration. doi:10.4073/csr.2006.4

38.

Obama

(The White House). (2014, 6 25). Empowering our children by bridging the word gap (video file). Retrieved from https://www.youtube.com/watch?v=NhC3n7oUm9U

39.

O'Brien

R.W.

D'Elio

M.A.

Vaden-Kiernan

Magee

Younoszai

Keane

M.J.

… Hailey

(2002). A descriptive study of Head Start families: FACES technical report I. Washington, DC: U.S. Department of Health and Human Services.

40.

Raudenbush

S. W.

Bryk

A. S.

(2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Sage.

41.

Semrud-Clikeman

M. S.

(2007). Social competence in children (pp. 1–9). New York, NY: Springer. doi:10.1007/978-0-387-71366-3_1

42.

Thomas

Zimmer-Gembeck

M. J.

(2007). Behavioral outcomes of parent-child interaction therapy and Triple-P-Positive Parenting Program: A review and meta-analysis. Journal of Abnormal Child Psychology, 35, 475–495. doi:10.1007/s10802-007-9104-9

43.

Vahedi

Farokhi

Farajian

(2012). Social competence and behavior problems in pre-school children. Iranian Journal of Psychiatry, 7, 126–134.

44.

Valentino

Comas

Nuttall

A. K.

(2014). Child maltreatment and early mother-child interactions. In Narvaez

Valentino

Fuentes

McKenna

Gray

(Eds.), Ancestral landscapes in human evolution (pp. 265–272). New York, NY: Oxford University Press.

45.

Wakschlag

L. S.

Briggs-Gowan

M. J.

Hill

Danis

Leventhal

B. L.

Keenan

… Carter

A. S.

(2008). Observational assessment of preschool disruptive behavior, Part I: Reliability of the disruptive behavior diagnostic observation schedule (DB-DOS). Journal of the American Academy of Child & Adolescent Psychiatry, 47, 623–631. doi:10.1097/CHI.0b013e31816c5c10

46.

Webster-Stratton

Reid

Beauchaine

(2011). Combining parent and child training for young children with ADHD. Journal of Clinical Child and Adolescent Psychology, 40, 191–203. doi:10.1080/15374416.2011.546044

47.

West

Denton

Germino-Hausken

(2000). America's kindergartners: Findings from the Early Childhood Longitudinal Study, kindergarten class of 1998-99, fall 1998 (NCES 2000-070). Washington, DC: U.S. Department of Education, NCES. Retrieved from http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2000070

48.

Yates

Ostrosky

M. M.

Cheatham

G. A.

Fettig

Shaffer

Santos

R. M.

(2008). Research synthesis on screening and assessing social-emotional competence. The Center on the Social and Emotional Foundations of Learning (CSEFL). Retrieved April 25, 2015, from http://csefel.vanderbilt.edu/documents/rs_screening_assessment.pdf

49.

Zimmer-Gembeck

M. J.

Thomas

(2010). Parents, parenting, and toddler adaptation: Evidence from a national, longitudinal study of Australian children. Infant Behavior and Development, 33, 518–529. doi:10.1016/j.infbeh.2010.07.004