Childhood Social Skills as Predictors of Middle School Academic Adjustment

Abstract

The present study used multiple regression analyses to examine the relationships between fifth-grade social skills and eighth-grade academic achievement. Data were drawn from the Early Childhood Longitudinal Study–Kindergarten Class of 1998-1999 (ECLS-K). Results indicated no relationship between positive or negative social behavior in fifth grade and academic achievement or teacher-rated academic skills in eighth grade. However, consistent with previous studies, fifth-grade approaches to learning were found to be positive predictors of both academic achievement and teacher-rated academic skills in eighth grade. In addition, these results suggest that socioeconomic status plays a significant and potentially unexplored avenue for understanding these outcomes. These results further illuminate the way behaviors in elementary school relate to academic adjustment to middle school.

Keywords

academic achievement prosocial behavior academic/school transitions social competence

Numerous educational policy statements (e.g., Adler, 1982; Boyer, 1983) suggest that the promotion of successful social skill development is a valued educational objective of the U.S. public schools (Wentzel, 1991). Significant effort has been expended to develop state academic standards for instruction, but less work has been devoted to developing and implementing standards and objectives in the social-emotional domain (Delpit, 2005; Wentzel 1991). Not every child starts school at the same level of social-emotional development. These differences in school readiness have been linked to disparities in academic achievement, high school graduation rates, and long-term employment opportunities (Bierman, Domitrovich, & Darling, 2007). One potential way to address these inequalities is to provide explicit instruction in social skills. Evidence suggests that students’ mastery of social skills is associated with better school performance (Durlak, Weissberg, Dymnicki, Taylor, & Schellinger, 2011; January, Casey, & Paulson, 2011).

Recent research has demonstrated that social skills interventions (e.g., training the skills of managing emotions, setting positive goals, taking another’s perspective, making informed decisions, and maintaining positive relationships) are most likely to succeed when they are applied during a developmental transition period (January et al., 2011). Most social skill instruction is aimed at the period during the preschool and primary years. Considerably less research has examined the efficacy of social skill interventions during another important developmental transition period, early adolescence. January et al. (2011) suggested a relationship between prosocial behavior and successful adjustment to middle school but noted that more research is needed. An understanding of the social skills that support adjustment to middle school would allow for the development of curricula that target such skills. As such, the purpose of this study was to examine if social skills at the end of elementary school relate to a student’s academic adjustment to the demands of the middle school environment.

Adjustment to Middle School

Although there currently is not a universal definition of adjustment to middle school, researchers have demonstrated that students who successfully transition to middle school do so through their relation to their peers (i.e., social adjustment), their academics (i.e., academic adjustment), and the overall state of their mental health (Farmer, Hall, Weiss, Petrin, Meece, & Moohr, 2011; McHale et al., 2005; Ryan & Shim, 2008). A student who is struggling to adjust to middle school often experiences peer rejection, grade decline, or lowered self-esteem (Farmer, Hall, Weiss, Petrin, Meece, & Moohr, 2011; McHale et al., 2005; Ryan & Shim, 2008). Students appear to adjust to middle school by developing a new set of social skills (January et al., 2011).

Peer rejection and low self-esteem are correlated with lower levels of academic achievement (Bellmore, 2011). Similarly, differences in middle school adjustment can have long-term effects on academic achievement and peer relations (Bellmore, 2011; Rueger, Malecki, & Demaray, 2011). Many students are at risk of victimization during the middle school years, and such victimization puts students at greater risk of experiencing later difficulties in academic and psychological functioning (Rueger et al., 2011). Thus, academic achievement can serve as one important indicator of a student’s adjustment to middle school. As such, this study focused on academic adjustment as an indicator of a student’s overall adjustment to middle school. In addition, the transition to middle school offers a critical opportunity to effect these long-term trajectories (Rueger et al., 2011).

Middle School and Negative Adjustment

The transition to middle school in the United States is associated with negative outcomes (Juvonen, Le, Kaganoff, Augustine, & Constant, 2004). Middle school frequently marks a period of academic decline and social maladjustment (Ryan & Shim, 2008). Rockoff and Lockwood (2010) performed a longitudinal study of New York City public middle schools and found that academic achievement declined during the specific year that students moved to a middle school. In addition, academic achievement continued to decline throughout middle school. It is not only academic achievement that declines during the middle school years but also motivation and social adjustment (Lepper, Corpus, & Iyengar, 2005; Ryan & Shim, 2008). One longitudinal study of rural adolescents found that their sense of belonging decreased at the onset of middle school and continued to decrease from sixth to 12th grades (Witherspoon & Ennett, 2011).

Decreases in achievement during middle school affect some groups more than others. A gap exists in achievement between students from homes with fewer economic resources and those from more affluent families (Balfanz & Byrnes, 2006). Until fourth grade, the achievement gap is relatively small, but between fourth grade and eighth grade, the gap in academic achievement widens significantly (Balfanz & Byrnes, 2006). Delpit (2005) suggested that the implicit social expectations within the classroom are often the social norms of individuals from higher socioeconomic status (SES), and teachers are often unaware of, or reluctant to explicitly communicate, these expectations. Students of lower SES attending schools that implicitly reinforce these expectations are placed at a disadvantage because teachers will reward and punish behaviors that have not been explicitly taught (Delpit, 2005). Wentzel (1991) concluded that, despite the fact children cite social-moral norms as the most important classroom rules, teachers spend very little or no time instructing students on these norms.

Adjustment During Early Adolescence

Adjustment to a new environment often requires both the acquisition of new skills and the adaptation of established skills to new situations. The transition to middle school often involves moving from a smaller building with established social relationships to a larger building with unfamiliar people and a new social context. This can lead to a shuffling of peer relationships (January et al., 2011). Adolescence is also a time when youth begin to assert their independence from adults and develop their identity in relation to their peers instead (Farmer, Hamm, Leung, Lambert, & Gravelle, 2011). These peer relationships assume even greater importance as early adolescents progress into high school (Rueger et al., 2011). It is likely that students with highly developed prosocial skills would more easily adapt to the middle school environment.

In addition to changes in social relationships, adolescents undergo major changes in physical and cognitive development during the middle school years. These changes often occur at differing developmental rates (Vawter, 2010). The structure of the brain undergoes considerable change during adolescence, and researchers have suggested that these changes have significant impacts on behavior (Vawter, 2010). Changes in physical maturity, mental maturity, and social relationships all have the possibility of impacting a student’s academic adjustment to middle school.

Contribution of Late Elementary Social and Academic Behavior

Differences in social skills and approaches to learning may explain the divergent trajectories of students entering middle school. Student behavior at the end of elementary school has been shown to predict certain aspects of adjustment (e.g., Bierman et al., 2008; Montague, Enders, Cavendish, & Castro, 2011). Bongers, Koot, van der Ende, and Verhulst (2004) found that teacher ratings of behavioral problems in elementary school are predictors of behavioral problems in adolescence. A longitudinal study conducted by Montague et al. (2011) demonstrated that teacher ratings of low academic competence during elementary school were associated with higher ratings of learning problems during high school. An additional finding was that an individual achievement test administered at the end of elementary school predicted reading and math achievement during high school (Montague et al., 2011).

Similarly, Wentzel (1993) reported that the ability to engage in certain social skills (e.g., managing emotions, setting positive goals, taking another’s perspective, making informed decisions, and maintaining positive relationships) can be predictive of future academic achievement. Social skills have been broadly defined as specific behaviors that lead to desirable social outcomes for the possessor (Whitcomb & Merrell, 2013). Approaches to learning represent a set of academically related attitudes and behaviors that predict scores from both standardized achievement measures and teacher assignment (Schaefer & McDermott, 1999; Trigwell, Ashwin, & Millan, 2013). Behaviors concomitant with approaches to learning include regulating problem solving, attending to tasks, and pursuing goals. According to January et al. (2011), individuals capable of accurately describing their relationships to their peers, their teachers, and their community will have better peer relations, higher self-esteem, and greater academic achievement.

Prior research (e.g., Caprara et al., 2000; Wentzel, 1993) has demonstrated that social skills in late elementary school have a predictive relationship with academic achievement. However, few studies have used a large and nationally representative data set to examine the relationships between late elementary school behavior and academic achievement. Further research is necessary to validate the relationship between social skills and academic achievement and to determine the social skills and specific behaviors that are associated with positive adjustment.

Rationale and Hypotheses

For many youth, middle school marks the beginning of a downward trend in academic adjustment (Ryan & Shim, 2008). One possible explanation for this decline is that students lack the requisite social skills to adjust to the demands of the middle school environment, which they perceive as foreign and unkind (January et al., 2011; Juvonen et al., 2004). Delpit (2005) hypothesized that schools and teachers lack the knowledge or, possibly, the will to address deficits in social skills. The identification of social skills associated with positive academic adjustment may aid in the development of social skill curricula that accurately target the needs of an early adolescent entering the middle school environment and, thereby, positively impact adjustment of students entering middle school. As such, the purpose of this study is to examine the relationships between specific social skills exhibited by students at the end of elementary school and their subsequent adjustment to middle school. Specifically, the present study will address the following research question:

Research Question 1: Are prior social skills related to positive academic adjustment in middle school?

The theoretical rationale for this study is that students who have greater social skill proficiency at the end of elementary school will be better able to adjust to middle school. Students who exhibit lower social skill proficiency at the end of elementary school may be challenged by the new demands of the middle school social landscape and show lower rates of academic achievement as a result. Students with lower social skill proficiency at the outset of middle school would be more likely to grasp the importance of developing social skills such as managing emotions, setting goals, and making responsible social decisions.

Based on the findings of previous studies (Durlak et al., 2011; January et al., 2011; Wentzel, 1993), social skills (e.g., the skills of managing emotions, setting positive goals, taking another’s perspective, making informed decisions, and maintaining positive relationships) were hypothesized to have a positive relationship with indicators of academic adjustment to middle school. Specifically, these indicators include academic achievement and teacher ratings of academic skills.

Method

Participants

Data for this study were drawn from the Early Childhood Longitudinal Study–Kindergarten Class of 1998-1999 (ECLS-K). The ECLS-K was a longitudinal study conducted by the National Center for Education Statistics (NCES) that followed a nationally representative cohort of 21,260 children from kindergarten through eighth grade. The primary objectives of the ECLS-K were to provide information on student achievement and developmental status in the elementary and middle school years, the strengths and weaknesses of U.S. kindergarten programs, and the relationships between home life and school experience as children move from elementary school to middle school. Data were collected in seven waves, twice during kindergarten, twice during first grade, and once during third, fifth, and eighth grades (The first through fifth waves were not used in the present study because they were not relevant to the research questions). The data in the sixth wave were collected in the spring of 2004 when approximately 90% of the sampled children were in the fifth grade (Pollack, Najarian, Rock, & Atkins-Burnett, 2005). The data in the seventh wave were collected in the spring of 2007 when approximately 89% of the sampled children were in the eighth grade (Najarian, Pollack, & Sorongon, 2009). Additional information on the participants can be found in the Eighth Grade User’s Manual for the ECLS-K (Tourangeau et al., 2009).

The sample of children from the seventh wave of data collection included 9,725 respondents. Due to participant attrition over time, demographic characteristics from the eighth-grade data represent most but not all eighth graders of the sample in the 2006-2007 academic year (Tourangeau et al., 2009). Examples of subgroups of eighth graders not represented in the data include children who immigrated to the United States after first grade or children who were home-schooled until first grade. However, given that the size of the sample, the demographic characteristics of the participants are very similar to the characteristics of the population of eighth graders during the 2006-2007 academic year.

Fifty percent of the sample was male, which matches the distribution of the general student population in 2009 (U.S. Census Bureau, 2010). The majority and minority status of the sample was also similar to that of the 2009 U.S. Census Projections. Sixty-two percent of the sample fell in the category of “Majority Status,” and 38.5% fell in the “Minority Status” category. Nineteen percent of the sample lived in the Northeast, 28% lived in the Midwest, 33% lived in the South, and 21% lived in the West. The disability status for the sample is one of the few variables that displayed much difference from the 2009 U.S. Census. Children with a disability comprised 7.9% of the sample, whereas children with a disability make up 13.1% of the public school population (U.S. Census Bureau, 2010). Further information regarding the SES level, area classification, and school type for the children from the sample can be found in Table 1.

Table 1.

Demographic Characteristics of Participants (N = 9,225).

Characteristic	n	%	% U.S. population
Male	4,627	50.2	49.2
Race/Ethnicity
Minority status	3,553	38.5	34.9
Majority status	5,663	61.5	65.1
Socioeconomic Status (SES) Level
First quintile	1,414	16.1
Second quintile	1,691	35.2
Third quintile	1,759	20.0
Fourth quintile	1,805	20.5
Fifth quintile	2,140	24.3
Region
Northeast	1,706	18.5	20.9
Midwest	2,549	27.8	21.7
South	2,998	32.6	34.1
West	1,939	21.0	23.3
Area
Large/mid-size city	2,755	29.9
Large/mid-size suburb or large town	3,345	36.3
Small town and rural	2,160	23.4
School Type
Public school	7,596	82.3	88.7
Private school	1,571	17.0	11.3
Disability Status
Child with disability	732	7.9	13.1
Child without disability	8,493	92.1	86.9

Note. Characteristics of the Wave 1 sample. SES = socioeconomic status.

Measures

The ECLS-K employed multiple methods and sources to study children’s early school experiences. These methods included direct assessments of participating children, interviews with parents, and data collection from teachers. The information gathered from these unique perspectives was used to measure the home life and school experiences of students from school entry to the end of middle school.

General education teacher questionnaire

Teacher questionnaires were used to obtain information about the sampled child’s social status relative to peers within the classroom and content-relevant skills. The approaches to learning, peer relations, and positive and negative social behavior ratings were all drawn from the general education teacher questionnaires.

Three subject-matter questionnaires were developed for the areas of reading, mathematics, and science in fifth grade. However, only two subject-matter questionnaires were completed for each sampled child in the study. Every child’s English teacher completed a questionnaire, and either the mathematics teacher or the science teacher completed the second questionnaire. The questionnaires were all organized in the same format, and began with the same series of questions regarding the child’s social skills and general class performance. Examples of questions related to social skills included, “Does this student seem to relate well to other students in your class?” and “Is this student exceptionally passive or withdrawn in your class?”

Questions regarding the child’s academic skills were based on the Academic Rating Scale (ARS). The sampled child’s skills were rated on a 5-point Likert-type scale ranging from 1 (poor) to 5 (outstanding). If a skill was not part of the curriculum, or if the teacher had not observed the skill, then the skill could be coded “not applicable or not observed.” Examples of skills included on the English questionnaire are “Ability to express analytical or critical thinking” and “Ability to gather and use information for research purposes.”

Cronbach’s coefficient alpha for the teacher rating scales was adequate and ranged from .93 to .96 (Najarian et al., 2009). Correlations among the teacher rating scales and direct cognitive assessment were examined for evidence of convergent and discriminant validity. Correlations among teacher ratings of skills and direct cognitive test scores for English, math, and science were .58, .57, and .51, respectively.

Approaches to learning scale

The Approaches to Learning scale is drawn from the ECLS-K Teacher Social Rating Scale (T-SRS). It has six items that assess a child’s attentiveness, task persistence, eagerness to learn, and learning independence. The split-half reliability for the approaches to learning scale was adequate (.91; Pollack et al., 2005).

Positive social behavior

Positive Social Behavior combines two scales, the Self-Control scale and the Interpersonal Skills scale. The Self-Control scale has four items that assess the student’s ability to control her or his temper, accept peer ideas, and respond appropriately to pressure. The Interpersonal Skills scale has five items that assess the child’s empathy and ability to build and maintain appropriate relationships with peers. Split-half reliability estimates were adequate (.79 for Self-Control and .88 for Interpersonal Skills; Pollack et al., 2005).

Negative social behavior

Negative Social Behavior combines the Externalizing Problem Behaviors scale and the Internalizing Problem Behaviors scale. The Externalizing Problem Behaviors scale has six items that assess the students’ frequency in engaging in arguments, impulsivity, and physical aggression. The Internalizing Problem Behaviors scale has four items that assess the appearance of behaviors relating to anxiety, low self-esteem, and sadness. Split-half reliability estimates were adequate (.89 for Externalizing and .77 for Internalizing; Pollack et al., 2005).

Reading

The ECLS-K reading battery used for fifth grade was adapted from the 1992 and 1994 National Assessment of Educational Progress (NAEP) Reading Frameworks, and the reading battery used for eighth grade was adapted from the 1992-2007 NAEP Reading Frameworks. The NAEP framework emphasizes four reading comprehension skills: understanding a text, interpreting a text, responding to a text, and critiquing a text. The eighth-grade reading battery covered the same concepts as the fifth-grade battery; however, the questions were adjusted to be appropriate for older students. In each battery, 10 initial items were used to route the child to a second-stage form that consisted of three or four reading passages, with each passage including three to nine associated questions. Items were scored on an Item Response Theory (IRT) scale. The reliabilities of theta for the reading assessments were adequate (.93 in the sixth wave and .87 in the seventh wave). Validity of the scores was supported by correlation of IRT theta scores between rounds of data collection. The correlation between the sixth and seventh waves of data collection was.79 (Najarian et al., 2009).

Mathematics

The mathematics battery used in the ECLS-K was developed from the 2005 NAEP Mathematics Framework for the eighth grade. The content of the mathematics battery included number sense and operations, measurement, geometry and spatial sense, statistics and probability, and algebra. Items for each of these content areas were crafted to measure basic to advanced mastery. The mathematics test began with a routing assessment of 10 items. The routing test was followed by one of two second-stage forms consisting of 20 items on each form. Items were scored on an IRT-scale. The reliability coefficients for the mathematics assessments were adequate (Wave 6 = .95 and Wave 7 = .92). Correlation between the direct cognitive measures supported the validity of the scores with a coefficient of .73 for the relationship between reading and mathematics. In addition, validity was supported by correlation between the direct and indirect assessments of mathematics achievement with coefficients ranging from .49 to .57 (Najarian et al., 2009).

Science

The science battery used in the ECLS-K was developed from the 1996 NAEP Science Framework and the results from a NAEP 2000 survey on middle school and high school science courses. The content of the science battery covered the topics of earth science, physical science, and life science. Similar to the mathematics assessment, the science test began with a routing assessment of 10 items. The routing test was followed by one of two second-stage forms consisting of 20 items on each form. Items were scored on an IRT-scale. The reliabilities for the science assessments were adequate (Wave 6 = .87 and Wave 7 = .84). The reliabilities for the science assessment are lower than for mathematics and reading, and the authors attributed the lower reliability scores to a greater diversity of content in the science assessment. Validity was examined via correlations between the reading and science measures (r = .77) and between the direct and indirect assessments of science achievement (r = .51; Najarian et al., 2009).

Procedures

Participants were assessed during the spring of their fifth- and eighth-grade years. During the fifth-grade year, the direct child assessments were administered individually in a computer-based format. During the eighth-grade year, the direct child assessments were administered to a group of students in a pencil and paper format. However, the structure and content of the computer-based assessment were not changed when converted to pencil and paper format. Information provided by the teachers was collected through a questionnaire. The information provided by parents was collected via an interview. The teacher questionnaire was organized into three sections that gathered information on the child’s social skills, the other children in the classroom, and the instructional practices in the classroom. The direct child assessments and the student and teacher questionnaires were conducted during the spring after obtaining parental consent. Further information on the data collection procedures can be found in the ECLS-K user’s manual (Tourangeau et al., 2009).

Data Analysis

The ECLS-K was intended to be a longitudinal study that was nationally representative. However, issues such as non-response, attrition, and immigration after first grade could create biases within the data. The ECLS-K recommends the use of sampling weights based on the child-level data. Consequently, the sampling weights for the sixth and seventh waves of data were selected because all analyses examined relationships between the sixth and seventh waves. Sampling weights were employed in all analyses.

Hierarchical multiple regression analyses were used to evaluate the relationships between fifth-grade social skills and eighth-grade middle school achievement. The decisions regarding variable entry were made based on the studies of Wentzel (1993) and Caprara et al. (2000). Specifically, these two studies used academic behaviors and peer relationships as covariates. Social skills were entered into the equation last. In the present study, approaches to learning was entered immediately following the demographic information because it represented academic behaviors. Then, following the pattern set out by Wentzel and Caprara, students’ ability to relate to peers was entered. Positive and negative social behavior were entered last.

The specific regression equation used to predict each of the eighth-grade outcomes was as follows:

\hat{Y} = b_{0} + b_{1} (PREV) + b_{2} (SES) + b_{3} (RACE) + b_{4} (SEX) + b_{5} (ATL) + b_{6} (TDQ) + b_{7} (PSB) + b_{8} (NSB) + e,

where Y was the outcome variable that the equation was predicting; PREV was the prior measure of the outcome variable; SES was family socioeconomic status; RACE was student race; SEX was student gender identity; ATL was the teachers’ rating of the students’ academic behaviors; TDQ was the teachers’ rating of how well students related with peers in the classroom; PSB were the positive social behaviors exhibited at the end of fifth grade as described by the teachers in the general education teacher questionnaire; and NSB were the negative social behaviors exhibited at the end of fifth grade as described by the teachers. An alpha level of .05 was used to test the statistical significance of the predictors in the model.

Results

Assumptions

Data were examined to determine if they met assumptions of normality prior to conducting the proposed analyses. Standardized residuals were plotted against predicted values, and outcome and predictor variables were plotted to test linearity and homoscedasticity. Examination of the plots indicated no curves and data scattered evenly around zero. Thus, the assumption of linearity was met. A scatterplot of the residuals against each of the predictor variables demonstrates that the assumption of independence of error is met. Histograms and p-p plots were used to test the normality of the residuals. Examination of the obtained histograms revealed them to have a normal, bell-curve shape, and the p-p plots show straight lines.

Guidelines for assessing multicollinearity provided by Bowerman and O’Connell (1990) and Myers (1990) indicate that when the largest variance inflation factor (VIF) exceeds 10, or when the average VIF is substantially greater than 1, there is cause for concern. Neither condition was met in the current data. In addition, Menard (1995) suggested that tolerance values below 0.1 indicate a serious problem, and tolerance values below 0.2 indicate potential problems. All tolerance values examined were greater than 0.2. Thus, based on these criteria, the data did not demonstrate multicollinearity.

The data met assumptions of normality and, as a result, were considered appropriate for analysis and interpretation.

Predicting Achievement on Tests of Academic Skills

Bivariate correlations were calculated between the outcome and predictor variables. The magnitude of the correlations ranged from .01 to .86. The value for the correlation between the math achievement score from fifth grade and the math achievement score from eighth grade was the highest at .86. Table 2 presents the results of the bivariate correlations between the outcome and predictor variables. Table 3 presents results of the multiple regression analyses of fifth-grade positive and negative social behavior predicting eighth-grade academic achievement as measured by standardized assessment of reading, math, and science with and without the use of a prior measure as a covariate.

Table 2.

Pearson Correlations Between All Tested Variables.

Characteristic	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.	13.	14.	15.	16.	17.	18.
1. Fifth-grade reading	1	.729	.759	−.055	.004	.208	.106	.135	.301	.044	.052	.729	.678	.717	.401	.450	.379	.328
2. Fifth-grade math		1	.747	−.038	.007	.202	.109	.128	.289	.044	.059	.631	.815	.728	.358	.401	.420	.312
3. Fifth-grade science			1	−.032	.027	.167	.095	.118	.251	.044	.054	.658	.675	.786	.357	.382	.357	.312
4. Fifth-grade externalizing problems				1	.526	.484	.419	.357	.162	.003	.010	−.064	−.053	−.046	−.039	−.061	−.036	−.034
5. Fifth-grade internalizing problems					1	.426	.399	.397	.141	−.001	.017	−.001	.001	.013	.014	.005	.000	.024
6. Fifth-grade approach to learning						1	.655	.578	.359	.023	.034	.200	.203	.180	.147	.186	.179	.178
7. Fifth-grade self-control							1	.531	.292	.015	.028	.099	.104	.093	.078	.101	.091	.116
8. Fifth-grade interpersonal skills								1	.234	.016	.026	.130	.117	.112	.121	.122	.126	.124
9. Fifth-grade literacy ARS									1	.030	.063	.253	.271	.251	.204	.221	.223	.169
10. Fifth-grade math ARS										1	−.889	.034	.034	.029	.026	.035	.215	.000
11. Fifth-grade science ARS											1	.048	.055	.057	.044	.043	.000	.101
12. Eighth-grade reading												1	.678	.706	.376	.423	.350	.329
13. Eighth-grade math													1	.754	.349	.395	.419	.309
14. Eighth-grade science														1	.356	.388	.395	.315
15. Eighth-grade oral ARS															1	.681	.355	.298
16. Eighth-grade written ARS																1	.390	.329
17. Eighth-grade math ARS																	1	.000
18. Eighth-grade science ARS																		1

Note. ARS = Academic Rating Scale.

Table 3.

Standardized Regression Coefficients in Predicting Eighth Grade Standardized Test Scores (IRT Scale) With and Without Prior Measure.

Predictor variable	Eighth-grade reading achievement without prior with prior				Eighth-grade math achievement without prior with prior				Eighth-grade science achievement without prior with prior
Predictor variable	β	ΔR²	β	ΔR²	β	ΔR²	β	ΔR²	β	ΔR²	β	ΔR²
Prior measure	—	—	.679	.622	—	—	.810	.742	—	—	.713	.664
Background variables
SES	.355	.228	.104	.011	.341	.206	.071	.005	.337	.230	.095	.010
Ethnicity	.146	.019	.051	.002	.117	.013	.001	.000	.224	.048	.048	.001
Gender	.013	.011	.019	.002	−.122	.001	.020	.002	−.166	.006	−.034	.000
Teacher-rated approaches to learning	.306	.083	.075	.008	.326	.087	.028	.003	.259	.069	.081	.009
Relates well to peers	−.027	.001	−.014	.000	−.015	.000	−.003	.000	−.015	.000	−.013	.000
Teacher-rated positive social behavior		.001		.000		.001		.001		.000		.000
Self-control	−.050		−.011		−.045		.013		−.038		−.004
Interpersonal	−.051		−.010		−.062		−.001		−.042		−.011
Teacher-rated negative social behavior		.006		.002		.006		.001		.007		.001
Externalizing	−.074		−.051		−.061		−.035		−.095		−.045
Internalizing	−.065		−.021		−.072		−.005		−.053		−.023

Note. Total sample (N = 9,225). SES = socioeconomic status.

All values significant at p < .001.

Approaches to learning (i.e., attention regulation, persistence, working in groups) in fifth grade was positively related with reading in eighth grade, explaining 8.3% of the variance. Teacher-rated peer relationships, positive social behavior, and negative social behavior were all negatively related with reading achievement in eighth grade. However, the strength of these relationships was weak. SES was a positive predictor of eighth-grade reading achievement indicating that children with higher SES scored higher in reading achievement in eighth grade. Ethnicity and gender were both predictors of eighth-grade reading skills, but they each explained only 1% of the variance. When reading achievement was entered as the first predictor in the regression equation, it accounted for the majority of the variance (Table 3). SES explained a small portion of the variance beyond the prior reading measure. The relationships between the remaining predictor variables and eighth-grade reading achievement were significant, but the strength of each relationship was so small as to be negligible (ΔR² < .01).

The results for mathematics and science demonstrated the same pattern of the relationships as those between social behavior and reading. SES had the strongest relationship with eighth-grade math and eighth-grade science achievement, respectively. The next largest predictor in both equations was approaches to learning. The relationship between fifth-grade approaches to learning and eighth-grade math achievement was positive (ΔR² = .087), and the relationship between fifth-grade approaches to learning and eighth grade science achievement was also positive (ΔR² = .069). Prior math achievement and prior science achievement explained the majority of the variance when entered into the respective equations (Table 3).

Predicting Teacher-Rated Academic Skills

A second set of regression analyses were run with the same predictor variables applied to predict teacher ratings of academic skills (ARS scores) instead of performance on standardized tests of achievement. Table 4 presents results of the multiple regression analyses of fifth-grade positive and negative social behavior predicting eighth-grade oral literacy, written literacy, math, and science ARS score, with and without the use of a prior measure as a covariate. The regression analyses were hierarchical, and the variables were added step by step. As each variable was added, a change in R² was calculated.

Table 4.

Standardized Regression Coefficients in Predicting Eighth Grade Teacher-Rated ARS Scores With and Without Prior Measure.

Predictor variable	Eighth-grade oral ARS score without prior with prior				Eighth-grade writing ARS score without prior with prior				Eighth-grade math ARS score without prior with prior				8th grade science ARS score without prior with prior
Predictor variable	β	ΔR²	β	ΔR²	β	ΔR²	β	ΔR²	β	ΔR²	β	ΔR²	β	ΔR²	β	ΔR²
Prior measure	—	—	.373	.274	—	—	.375	.332	—	—	.310	.235	—	—	.260	.211
Background variables
SES	.273	.131	.199	.042	.289	.152	.214	.047	.215	.108	.151	.035	.305	.171	.237	.072
Ethnicity	.039	.001	.040	.001	.033	.000	.033	.001	.070	.003	.076	.004	.085	.006	.073	.003
Gender	.035	.016	.022	.002	.102	.046	.090	.015	.013	.012	.041	.010	.029	.017	.050	.014
Approaches to learning	.309	.082	.102	.006	.386	.120	.176	.017	.446	.122	.271	.037	.289	.090	.189	.035
Relates well to peers	−.164	.026	−.152	.021	−.144	.020	−.132	.016	−.046	.002	−.057	.003	−.122	.014	−.129	.015
Positive social behavior		.005		.001		.004		.002		.007		.003		.005		.002
Self-control	−.087		−.031		−.088		−.029		−.013		−.002		−.133		−.072
Interpersonal	.012		−.012		−.031		−.055		−.123		−.090		.099		.076
Negative social behavior		.006		.002		.004		.001		.009		.003		.000		.000
Externalizing	.035		.027		−.015		−.022		.042		.012		−.033		−.001
Internalizing	−.087		−.048		−.072		−.032		−.106		−.067		−.002		.022

Note. Total sample (N = 9,225). ARS = Academic Rating Scale; SES = socioeconomic status.

All values significant at p < .001.

Approaches to learning in fifth grade and oral literacy score in eighth grade were positively related (ΔR² = .082). Teacher-rated peer relationships had a negative and significant relationship with oral literacy score (ΔR² = .026). Positive social behavior and negative social behavior were both predictors of oral literacy score in eighth grade, but the strength of these relationships was negligible. SES was a positive predictor of oral literacy score indicating that children with higher SES were rated higher on the oral literacy score in eighth grade. Ethnicity and gender were both significant predictors of eighth-grade oral literacy score, but they each explained only 1% of the variance. Prior reading score explained a large portion of the variance when entered into the equation (Table 4).

Fifth-grade teachers rated students on reading skills, whereas eighth-grade teachers were asked to rate students on oral literacy and written language skills. As a result, fifth-grade reading ARS score was used as a prior measure of both oral literacy ARS score and written literacy ARS score. SES explained a small portion of the variance (ΔR² = .042). Teacher-rated peer relationships and oral literacy score exhibited a negative predictive relationship. The relationships between the remaining predictor variables and eighth-grade oral literacy were significant, but the strength of each relationship was negligible.

The relationships between predictor variables and eighth-grade written language ratings follow a similar pattern to those found with oral literacy. Approaches to learning in fifth grade and written literacy score in eighth grade were positively related (ΔR² = .12). Teacher-rated peer relationships had a negative relationship with written literacy score (ΔR² = .02). SES was a positive predictor of eighth-grade written literacy score and accounted for the largest proportion of variance. Ethnicity was a predictor of eighth-grade written literacy score, but the strength of the relationship was negligible. Gender was a predictor of eighth-grade written literacy score, indicating that girls were rated higher than boys. When the fifth-grade reading ARS score was entered into the equation first, however, it accounted for the greatest proportion of the variance, and all other relationships were negligible.

The predictor variables demonstrated the same pattern of relationships with math and science ARS scores. Approaches to learning in fifth grade were positively related with both math (ΔR² = .12) and science (ΔR² = .09) scores in eighth grade. SES was a positive predictor for both eighth-grade math and science scores. Teacher ratings of math and science proficiency in fifth grade explained the largest proportion of the variance when entered into each equation (Table 4). SES and approaches to learning, however, still demonstrate a positive relationship with both the eighth-grade math (SES ΔR² = .035; approaches to learning ΔR² = .037) science scores (SES ΔR² = .072; approaches to learning ΔR² = .035).

Discussion

The purpose of this study was to examine the relationships between social skills exhibited at the end of elementary school and students’ academic adjustment to middle school. Contrary to predictions, the strength of the relationship between positive and negative social behavior in fifth grade and academic skills in eighth grade was negligible across all academic areas, regardless of the method of measurement (direct assessment or teacher ratings). Consistent with previous research (DiPerna, Lei, & Reid, 2007), fifth-grade approaches to learning was a positive predictor of both academic achievement and teacher-rated academic skills in eighth grade. However, when a prior measure of academic achievement was used as a covariate, the strength of the relationship between fifth-grade approaches to learning and eighth-grade academic achievement was made negligible. When a prior measure of teacher-rated academic skill was used as a covariate, the strength of the relationship between fifth-grade approaches to learning and eighth-grade teacher-rated math and science skill was reduced, and no relationship was found between fifth-grade approaches to learning and eighth-grade teacher-rated oral language and writing skill. Finally, SES was found to be a positive predictor across all subjects and measures over and above ethnicity and gender, but the relationship was negligible when a prior measure of achievement was entered into the equation first.

Social behavior was hypothesized to predict academic adjustment over and above SES, race, gender, academic behaviors, and quality of peer relations. The results of the current analyses did not support this hypothesis. The magnitude of the relationship between positive and negative social behavior and academic achievement was so small as to be negligible. In addition, the relationship between approaches to learning and academic achievement and teacher-rated academic skill was small. One possible explanation is that approaches to learning and positive social behavior share an underlying third construct, such as self-regulatory behavior. Both approaches to learning and positive social behavior contain elements of self-regulatory behaviors, such as working independently or ignoring distractions, within a broader construct. Thus, when approaches to learning is included as a predictor, the positive and negative social behavior variables do not explain anything above and beyond approaches to learning. This potential explanation is further supported by the fact that when approaches to learning was removed as a covariate, positive social behavior was found to be a positive predictor, and negative social behavior was found to be a negative predictor, of academic achievement in eighth grade. Furthermore, several studies (e.g., Fennis, 2011; Lopes, Salovey, Côté, Beers, & Petty, 2005; Lopes, Salovey, & Straus, 2003) have examined the relationships between self-regulatory behavior and prosocial behavior, and, although the labels assigned to these behaviors were different from study to study, each found self-regulatory behavior, similar to the behaviors described by the ECLS-K Approaches to Learning scale, to be predictive of prosocial behaviors.

The finding that fifth-grade approaches to learning is a positive predictor of later academic achievement and teacher-rated academic skills is consistent with prior research concerning academic achievement and social behavior (DiPerna et al., 2007). Also, consistent with this previous research was the finding that the magnitude of the relationship was small. It should be noted that the relationships were reduced and, in many instances, made negligible when the fifth-grade measure of the outcome variable was used as a covariate. However, findings across studies (e.g., Caprara et al., 2000; DiPerna et al., 2007; Wentzel, 1993) suggest that approaches to learning has a small positive relationship with academic achievement. DiPerna et al. (2007) drew a sample from the same data set as the one in this study (i.e., the ECLS-K) and found approaches to learning to be a positive predictor of academic achievement in the elementary grades. Specifically, DiPerna, Lei, and Reid found approaches to learning at kindergarten entry to be a positive predictor of growth in mathematics achievement through the end of third grade. The similarity in instrumentation may have contributed to similarity in findings.

The findings regarding positive and negative social behavior are inconsistent with Wentzel (1993) and Caprara et al. (2000). Both the Wentzel and Caprara et al. studies examined smaller data sets and used methods, such as sociometrics, with greater alignment to the research questions regarding the social behavior of the children in the study. Specifically, the study conducted by Wentzel used peer nominations to rank students’ positive and negative social behavior. Caprara et al. used a similar type of sociometric measure as the Wentzel study. In addition, both studies used a composite of grades assigned by teachers as the academic achievement variable. In contrast, the ECLS-K provides standardized achievement measures and teacher ratings of academic skills. Both the Caprara et al. and Wentzel studies used a cognitive measure as a covariate, whereas the current study used prior measures of achievement. Finally, the smaller sample in the Wentzel and Caprara studies may have created the opportunity for the use of more comprehensive, time-intensive measures, which would allow for greater alignment between the research questions and the constructs being measured. Conversely, the use of a smaller sample can lead to a reduction in external validity due to a less representative sample. The Caprara et al. study, for example, drew its sample from a single community near Rome, Italy, and the sample from the Wentzel study was drawn from a single middle school from a Midwestern community, whereas the ECLS-K sample was drawn from a representative national sample.

The relationships between approaches to learning and teacher ratings of academic skills were found to be stronger than the relationships between approaches to learning and academic achievement scores. The difference ranged from 2% to 4% of the variance explained, and this finding held true regardless of whether or not a prior measure was entered into the equation. The relationship between teacher ratings of oral language academic skills and approaches to learning was the only relationship smaller than its academic achievement counterpart (i.e., reading achievement scores), and the relationship was only slightly smaller regardless of the inclusion of a prior measure in the equation.

The finding that fifth-grade approaches to learning has the most robust relationship with teacher ratings of math and science skills in eighth grade is consistent with previous studies (DiPerna et al., 2007). The magnitude of the relationships was substantially reduced after a prior measure was introduced to the equation. The relationships between teacher ratings of math and science skills and approaches to learning were the only relationships that were maintained after a prior measure was included. DiPerna et al. (2007) examined relationships between social behavior and mathematics achievement using the same data set and found similar results.

The relatively stronger relationship between approaches to learning and the teacher ratings of academic skill may be explained by teachers’ inclusion of additional factors in assigning academic skill ratings. There is evidence to suggest that teachers consider other elements of a student’s work when assigning grades, such as effort and participation (Hamre & Pianta, 2001). Teachers may have included these aspects of classroom behavior when assigning the ratings of academic skill. Thus, the teacher ratings of academic skills involve elements that are missed in a strict assessment of academic achievement.

Examination of the differences in relationships between student SES and student minority status with academic achievement produced an additional outcome worthy of note. After the prior measure, student SES was consistently the strongest predictor of academic achievement across all subjects and teacher ratings. In contrast, student minority status demonstrated a weak or no relationship with academic achievement and teacher ratings of academic skill. The analyses were completed again while entering student minority status into the equation first, and student SES remained the strongest predictor of academic achievement. This finding contradicts a number of studies (e.g., Posselt, Jaquette, Bielby, & Bastedo, 2012) suggesting that the differences in academic achievement are a result of minority status. The findings from this study, consistent with the work of Sirin (2005), suggest that student socioeconomic status is the strongest predictor of academic achievement, over and above student minority status.

Limitations and Directions for Future Research

There are a number of limitations to this study. First, the study is non-experimental and all relationships are correlational. Another such limitation was that student grades were not collected. Evidence suggests that grades are often assigned based on not only student academic achievement but also their approach to learning and behavior in the classroom. Teachers frequently take additional factors, such as perseverance and class participation, into account when assigning grades (Hamre & Pianta, 2001). Subsequently, grades may have provided a better understanding of how the student had adjusted to middle school because the grades would have also represented the effort spent in the classroom or the behavior of the student. It should also be noted that teachers may have taken additional factors, such as perseverance or participation, into account when assigning the ratings of academic skill. A second limitation is that the teacher rating scales of social behavior were designed for a very large study. As a result, these measures were intentionally brief and may not have adequately sampled the breadth of the key constructs in this study. In addition, as the developers of the ECLS-K acknowledged, it is difficult to assess all of the curricula being taught across schools that have decentralized control of curricula (Rock & Pollack, 2002). Despite the robust psychometric properties of the achievement measures developed for the ECLS-K, those measures may not have included curricula specific to a region or a school. As a result, students with growth in these areas would have gone unmeasured.

The ECLS-K was, by design, a study that included many different measures from multiple respondents. As a result, many of the instruments used to measure key variables, such as the behavioral constructs, had to strike a balance between brevity and depth of coverage. Furthermore, the measures used to assess social behavior were selected by the authors of the ECLS-K and were not necessarily selected to answer the research questions of the present study. The social skill measures selected for the present study were the ones that were most closely aligned with the hypothesized positive social behavior. In similar fashion, the measures used to assess academic achievement have excellent psychometric properties for capturing academic achievement. However, Rock and Pollack (2002), in an article describing the development of the ECLS-K achievement measures, noted that determining the academic skills typically taught in an education system with decentralized control proved to be a challenge. The academic measures may have missed some of the broader aspects of the curricula taught in the transition across middle school.

Finally, the predictors and covariates for this study had to be selected from those available within the Early Childhood Longitudinal Study. In contrast, the Wentzel (1993) and Caprara et al. (2000) studies both used measures specific to their purposes. As a result, the Wentzel and Caprara et al. studies both had a strong alignment between the research questions and the constructs being measured, but the psychometric properties of the instruments were not as strong as those in the ECLS-K. In addition, the possibility of researching possible mediators or other potential predictors of academic adjustment in middle school is restricted by the variables available within the ECLS-K.

Several important directions for future research should be considered. First, the analyses conducted in this study should be repeated with other data sets to determine if the pattern of results is unique to the ECLS-K, or if different instrumentation or samples may yield different results. Second, possible mediators between early social behavior and later academic achievement should be considered. Wentzel found academic behavior and teacher’s preferences for students to be potential mediators between social behavior and grade point average (GPA). A structural equation model, such as the one used by Caprara et al. (2000) may further illuminate the relationships between social behavior, academic behavior, and later achievement. Finally, this study focused specifically on academic adjustment to middle school, but other indicators of adjustment to middle school such as self-esteem, peer acceptance, peer rejection, and well-being could be considered in future studies.

Conclusion

Results of the current study provide insights regarding the way behaviors in elementary school relate to academic adjustment to middle school. Prior research (e.g., Caprara et al., 2000; Wentzel, 1993) demonstrated a relationship between social skills in elementary school and academic achievement in middle school. The results of this study indicate that fifth-grade approaches to learning is a small positive predictor of both academic achievement and teacher-rated academic skills in eighth grade. The inclusion of a prior measure of achievement as a covariate, however, yielded no relationship between fifth-grade approaches to learning and eighth-grade academic achievement. The inclusion of a prior measure of teacher-rated academic skills yielded weak relationships between fifth-grade approaches to learning and eighth-grade teacher-rated academic skills in math and science and no relationship with eighth-grade teacher-rated academic skills in oral language and writing. However, considering these results in conjunction with similar studies (e.g., DiPerna et al., 2007; Wentzel, 1993), it is likely that approaches to learning and later academic achievement have a small positive relationship. Behaviors associated with approaches to learning, such as persistence and attention regulation, have demonstrated a positive relationship with academic achievement (DiPerna et al., 2007; Schaefer & McDermott, 1999). Furthermore, the finding of a small positive relationship between approaches to learning and later academic achievement in a large nationally representative data set suggests that behaviors associated with approaches to learning influence academic success across settings. The results indicated no relationship between social behavior in fifth grade and academic achievement in eighth grade. Researchers interested in developing interventions for the adolescent period may use the results of this study by focusing on the instruction of self-regulatory social skills associated with academic behaviors, such as diligence and goal-oriented behavior, because these skills as measured by the approaches to learning variable showed the strongest and most consistent relationship with later academic achievement and may be used to better predict academic achievement.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Gordon Emmett Hall is a doctoral candidate in the school psychology program at the Pennsylvania State University. He is most interested in social and emotional development and school-based intervention and prevention programs that promote positive academic outcomes. His adviser is Dr. James DiPerna.

James Clyde DiPerna is an associate professor in the Department of School Psychology, Special Education, and Educational Psychology. His research interests include academic and emotional competence, school-based prevention programs, and the use of technology to facilitate learning.

References

Adler

(1982). The Paideia proposal: An educational manifesto. NY: Macmillan.

Balfanz

Byrnes

(2006). Closing the mathematics achievement gap in high-poverty middle schools: Enablers and constraints. Journal of Education for Students Placed at Risk, 11, 143-159. Retrieved from http://www.csos.jhu.edu/pubs/EdWeek/JESPAR%20Closing%20Achievement%20Gap%20Re-Print.pdf

Bellmore

(2011). Peer rejection and unpopularity: Associations with GPAs across the transition to middle school. Journal of Educational Psychology, 103, 282-295. doi:10.1037/a0023312

Bierman

K. L.

Domitrovich

Darling

(2007). Early prevention initiatives. J. Johnson (Ed.), Approaches to early childhood education (5th ed., pp. 147-164). Columbia, OH: Pearson Merrill Prentice Hall.

Bierman

K. L.

Domitrovich

Nix

R. L.

Gest

S. D.

Welsh

J. A.

. . . Gill

(2008). Promoting academic and social-emotional school readiness: The head start REDI program. Child Development, 79, 802-1817. doi:10.1111/j.1467-8624.2008.01227.x

Bongers

I. L.

Koot

H. M.

van der Ende

Verhulst

F. C.

(2004). Developmental trajectories of externalizing behaviors in childhood and adolescence. Child Development, 75, 1523-1537. Retrieved from http://www.jstor.org.ezaccess.libraries.psu.edu/stable/pdfplus/3696498.pdf?acceptTC=true

Bowerman

B. L.

O’Connell

R. T.

(1990). Linear statistical models: An applied approach (pp. 106-129). Boston, MA: PWS-Kent.

Boyer

E. L.

(1983). High school: A report on secondary education in America (The Carnegie Foundation for the Advancement of Teaching). New York, NY: Harper & Row.

Caprara

G. V.

Barbaranelli

Pastorelli

Bandura

Zimbardo

P. G.

(2000). Prosocial foundations of children’s academic achievement. Psychological science, 11(4), 302-306. doi:10.1111/1467-9280.00260

10.

Delpit

(2005). Other people’s children. New York, NY: The New Press.

11.

DiPerna

J. C.

Lei

P. W.

Reid

E. E.

(2007). Kindergarten predictors of mathematical growth in the primary grades: An investigation using the Early Childhood Longitudinal Study–Kindergarten cohort. Journal of Educational Psychology, 99, 369. doi:10.1037/0022-0663.99.2.369

12.

Durlak

J. A.

Weissberg

R. P.

Dymnicki

A. B.

Taylor

R. D.

Schellinger

K. B.

(2011). The impact of enhancing students’ social and emotional learning: A meta-analysis of school-based universal interventions. Child Development, 82, 405-432. doi:10.1111/j.1467-8624.2010.01564.x

13.

Farmer

T. W.

Hall

C. M.

Weiss

M. P.

Petrin

R. A.

Meece

J. L.

Moohr

(2011). The school adjustment of rural adolescents with and without disabilities: Variable and person-centered approaches. Journal of Child and Family Studies, 20, 78-88. doi:10.1007/s10826-010-9379-2

14.

Farmer

T. W.

Hamm

J. V.

Leung

Lambert

Gravelle

(2011). Early adolescent peer ecologies in rural communities: Bullying in schools that do and do not have a transition during the middle grades. Journal of Youth and Adolescence, 40, 1106-1117. doi:10.1007/s10964-011-9684-0

15.

Fennis

B. M.

(2011). Can’t get over me: Ego depletion attenuates prosocial effects of perspective taking. European Journal of Social Psychology, 41, 580-585. doi:10.1002/ejsp.828

16.

Hamre

B. K.

Pianta

R. C.

(2001). Early teacher-child relationships and the trajectory of children’s school outcomes through eighth grade. Child Development, 72, 625-638. doi:10.1111/1467-8624.00301

17.

January

A. M.

Casey

R. J.

Paulson

(2011). A meta-analysis of classroom-wide interventions to build social skills: Do they work? School Psychology Review, 40, 242-256. Retrieved from http://search.proquest.com.ezaccess.libraries.psu.edu/docview/878153409/fulltextPDF?accountid=13158

18.

Juvonen

Kaganoff

Augustine

Constant

(2004). Focus on the wonder years: Challenges facing the American middle school. Santa Monica, CA: RAND Corporation.

19.

Lepper

M. R.

Corpus

J. H.

Iyengar

S. S.

(2005). Intrinsic and extrinsic motivational orientations in the classroom: Age differences and academic correlates. Journal of Educational Psychology, 97, 184-196. doi:10.1037/0022-0663.97.2.184

20.

Lopes

P. N.

Salovey

Côté

Beers

Petty

R. E.

(2005). Emotion regulation abilities and the quality of social interaction. Emotion, 5(1), 113-118. doi:10.1037/1528-3542.5.1.113

21.

Lopes

P. N.

Salovey

Straus

(2003). Emotional intelligence, personality, and the perceived quality of social relationships. Personality and individual Differences, 35, 641-658. doi:10.1016/S0191-8869(02)00242-8

22.

McHale

J. P.

Vinden

P. G.

Bush

Richer

Shaw

Smith

(2005). Patterns of personal and social adjustment among sport involved and noninvolved urban middle-school children. Sociology of Sport Journal, 22, 119-136. Retrieved from http://journals.humankinetics.com.ezaccess.libraries.psu.edu/AcuCustom/SiteName/Documents/DocumentItem/4860.pdf

23.

Menard

(1995). Applied logistic regression analysis (Sage University Paper Series on Quantitative Applications in the Social Sciences). Thousand Oaks, CA: Sage.

24.

Montague

Enders

Cavendish

Castro

(2011). Academic and behavioral trajectories for at-risk adolescents in urban schools. Behavioral Disorders, 36, 141-156. Retrieved from http://web.ebscohost.com.ezaccess.libraries.psu.edu/ehost/pdfviewer/pdfviewer?sid=32bce6d6-2fac-4b86-8703-e3bb0daf4b1b%40sessionmgr110&;vid=2&hid=104

25.

Myers

R. H.

(1990). Classical and modern regression with applications (Vol. 2). Belmont, CA: Duxbury Press.

26.

Najarian

Pollack

J. M.

Sorongon

A. G.

(2009). Early Childhood Longitudinal Study, Kindergarten Class of 1998-99 (ECLS-K), Psychometric Report for the Eighth Grade (NCES 2009-002). Washington, DC: National Center for Educational Statistics, Institute of Education Sciences, U.S. Department of Education. Retrieved from http://nces.ed.gov/pubs2009/2009002.pdf

27.

Pollack

J. M.

Najarian

Rock

D. A.

Atkins-Burnett

(2005). Early Childhood Longitudinal Study, Kindergarten Class of 1998-1999 (ECLS-K) Psychometric Report for the Fifth Grade. NCES 2006?036. Washington, DC: National Center for Educational Statistics, Institute of Education Sciences, U.S. Department of Education. Retrieved from http://eric.ed.gov/?id=ED489079

28.

Posselt

J. R.

Jaquette

Bielby

Bastedo

M. N.

(2012). Access without equity longitudinal analyses of institutional stratification by race and ethnicity, 1972-2004. American Educational Research Journal, 49(6), 1074-1111. doi:10.3102/0002831212439456

29.

Rock

D. A.

Pollack

J. M.

(2002, August). Early Childhood Longitudinal Study–Kindergarten Class of 1998-99 (ECLS-K): Psychometric Report for Kindergarten through First Grade (Working Paper Series). Retrieved from http://nces.ed.gov/pubs2002/200205.pdf

30.

Rockoff

J. E.

Lockwood

B. B.

(2010). Stuck in the middle: Impacts of grade configuration in public schools. Journal of Public Economics, 94, 1051-1061. doi:10.1016/j.jpubeco.2010.06.017

31.

Rueger

S. Y.

Malecki

C. K.

Demaray

M. K.

(2011). Stability of peer victimization in early adolescence: Effects of timing and duration. Journal of School Psychology, 49, 443-464. doi:10.1016/j.jsp.2011.04.005

32.

Ryan

A. M.

Shim

S. S.

(2008). An exploration of young adolescents’ social achievement goals and social adjustment in middle school. Journal of Educational Psychology, 100, 672-687. doi:10.1037/0022-0663.100.3.672

33.

Schaefer

B. A.

McDermott

P. A.

(1999). Learning behavior and intelligence as explanations for children’s scholastic achievement. Journal of School Psychology, 37(3), 299-313. doi:10.1016/S0022-4405(99)00007-2

34.

Sirin

S. R.

(2005). Socioeconomic status and academic achievement: A meta-analytic review of research. Review of Educational Research, 75, 417-453. doi:10.3102/00346543075003417

35.

Tourangeau

Nord

Sorongon

A. G.

Najarian

Hausken

E. G.

(2009). Early Childhood Longitudinal Study, Kindergarten Class of 1998-99 (ECLS-K) (Combined User’s Manual for the ECLS-K Eighth-Grade and K-8 Full Sample Data Files and Electronic Codebooks [NCES 2009-004]). Washington, DC: National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education. Retrieved from http://nces.ed.gov/ecls/data/ECLSK_K8_Manual_part1.pdf

36.

Trigwell

Ashwin

Millan

E. S.

(2013). Evoked prior learning experience and approach to learning as predictors of academic achievement. British Journal of Educational Psychology, 83(3), 363-378. doi:10.1111/j.2044-8279.2012.02066.x

37.

U.S. Census Bureau. (2010). Summary File 1 (SF 1). Available from http://factfinder.census.gov

38.

Vawter

(2010). Mining the middle school mind. The Education Digest, 75(5), 47-49. Retrieved from http://ezaccess.libraries.psu.edu/login?url=http://search.proquest.com.ezaccess.libraries.psu.edu/docview/89155800?accountid=13158

39.

Wentzel

K. R.

(1991). Social competence at school: Relation between social responsibility and academic achievement. Review of Educational Research, 61, 1-24. doi:10.3102/00346543061001001

40.

Wentzel

K. R.

(1993). Does being good make the grade? Social behavior and academic competence in middle school. Journal of Educational Psychology, 85, 357-364. doi:10.1037/0022-0663.85.2.357

41.

Whitcomb

Merrell

K. W.

(2013). Behavioral, social, and emotional assessment of children and adolescents. New York, NY: Routledge.

42.

Witherspoon

Ennett

(2011). Stability and change in rural youths’ educational outcomes through middle and high school years. Journal of Youth and Adolescence, 40, 1077-1090. doi:10.1007/s10964-010-9614-6