The influence of behavioral and emotional characteristics on academic achievement of middle school students: A growth modeling approach

Abstract

The purpose of this study was to examine the associations between behavioral and emotional characteristics and middle school student achievement across different grades based on a growth modeling approach. Using a total of 1,874 students, target predictor variables (i.e., attention, aggressiveness, behavioral control, social withdrawal, depression, self-esteem) and dependent variables (i.e., Korean language arts, mathematics) were extracted from a national and longitudinal data set, and four predictor models were formulated to examine the influence of behavioral/emotional characteristics on student growth trajectories. Results showed that (a) students' initial performance at seventh grade did not predict their over-time growth; and (b) self-esteem and behavioral control variables impacted on the seventh graders' achievement as well as their growth from the seventh to ninth grade. Based on the findings, practical implications and future research are discussed.

Keywords

achievement behavioral and emotional characteristics growth modelling middle school students

Substantial research has yielded strong evidence that students' behavioral and emotional characteristics in classrooms impact academic performance (Sektnan, McClelland, Acock, & Morrison, 2010; Spengler, Damian, & Roberts, 2018; Sun & Shek, 2012). This research has found that, to be successful in school, students are required to remain on-task and complete given tasks, which can lead to positive learning outcomes. Students are also expected to build positive relationships with their peers as well as their teachers, as their learning processes and outcomes are influenced by their relationships with peers and teachers (Bronfenbrenner & Morris, 1998; DiPrete, & Jennings, 2012; Sektnan et al., 2010; Stuhlman & Pianta, 2009). However, students' inability to attend to instruction and complete tasks, which is often rooted in their behavioral and emotional characteristics, negatively impacts their educational performance (Ansary & Luthar, 2009; Frazier, Youngstrom, & Naugle, 2007; Masten et al., 2005; Merrell & Tymms, 2001; van Lier et al., 2012). Similarly, poor relations with teachers and their peers adversely impact on behaviors, often resulting in misbehaviors or antisocial behaviors that inhibit positive learning outcomes (Lansford et al., 2002; Spira & Fischel, 2005). In this sense, the skills listed above (i.e., being on-task, completing given tasks, building relationships with others) commonly underlie behaviors that are appropriately shaped and presented in accordance with a given context and/or situation. Not only are these required skills for academic success in school, these are skills that our society expects individuals to possess. Unfortunately, not all students have such positive behavioral skills, and several internalizing and externalizing behaviors impede learning processes in the classroom (Tompson, 2009).

The terms internalizing and externalizing behaviors refer to the two types of major behaviors students and adolescents exhibit during their schooling as well as throughout adulthood (Achenbach, 1966, 1978). Externalizing behaviors include behaviors that are visible and/or observable in a certain environment, such as aggressiveness, impulsivity, physical aggressiveness and lack of self-control; internalizing problem behaviors are not easily visible, but they can be shown through the expression of negative feelings and emotional responses to a specific environment or behaviors such as fearfulness, physical complaints, and depression. The direction of each behavior type is different; externalizing behaviors are directed outward (i.e., toward a student's environment or others) and internalizing behaviors are directed inward (i.e., toward the himself/herself). Both types of behavior, however, can affect teaching-learning processes in school (e.g., classroom settings), particularly when a student has inappropriate levels and/or expression of each behavior type (Baker, Grant, & Morlock, 2008; Barriga et al., 2002). Furthermore, the delay and/or lack of social and personal skills due to internalizing and/or externalizing behaviors also influences the lives of students outside of school (Capros, Cetera, Ogden, & Rossett, 2002; Pierson, Carter, Lane, & Glaeser, 2008).

Behavioral and emotional factors and student achievement

It is well known that students' ability to focus on instruction and to complete given tasks can affect students' academic performance (Keller, 1983; Skinner, Wellborn, & Connell, 1990; Zimmerman, 1989). Such studies focusing on the relationship between emotional and behavioral characteristics and academic achievement support the idea that an individual's behavioral and emotional factors impact their attention, which in turn influences their learning process (Pekrun, 2006). These studies have a common underlying assumption: The behavioral issues students present are antecedent and the learning outcomes or academic achievement are consequent (Coie, 1996; Kellam et al., 1991). Under this hypothetical proposition, behavioral problems make it difficult for students to focus on their learning, which in turn leads to low academic achievement (Chu & Baker, 2015). Thus, according to this perspective, modifications or interventions to problematic behaviors can lead to improved learning outcomes.

Research has also shown that the influences of behavioral characteristics have been found in different academic areas (Langberg & Becker, 2012; McClelland, Morrison, & Holmes, 2000). Trzesniewski et al. (2006) found, for example, that being antisocial at age 5-years-old predicted reading skill in 7-year-olds, when controlling for the child's age 5-years-old cognitive ability. Their findings also showed that being antisocial at age 7-years-old was negatively associated with reading achievement at age 7-years-old, indicating that being antisocial impacted on student achievement both crosswise and longitudinally. Similarly, in a study conducted by Hagan-Burke et al. (2011), results show that externalizing problem behaviors of kindergarten students (average 5.44 years) were significantly associated with reading performance for phonemic sound matching and word identification. Also, externalizing problem behaviors and hyperactivity, along with higher internalizing behavior ratings, were strongly associated with low word identification.

Behaviors and emotions that tend to leave students behind academically often result in eligibility for special education services. Students who are characterized as having emotional and behavioral problems, for example, make up a growing number of school-age students in public schools (Forness, Kim & Walker, 2012). This growing number of students with emotional and behavioral problems is concerning. It is particularly due to the fact that students with emotional and behavioral problems have a much higher dropout percentage than their peers and a higher chance of being arrested in middle or high school (Gage, Josephs, & Lunde, 2012; Kauffman, Mock & Simpson, 2007). The cycle frequently occurs because these students are very likely to become frustrated with academics in school settings and to externalize or internalize their frustration, thus falling even further behind. Therefore, it is expected that an understanding of the associations between behavioral and emotional characteristics and student academic performance provides a potential key to disrupting the negative cycle addressed above.

Problem statement and research questions

The majority of previous studies on student behaviors and their emotional characteristics have focused on the roles and effects of their behaviors on life success through occupational status and/or academic performance in a cross-sectional way (e.g., correlations among behaviors and academic performance; Bean, Bush, McKenry, & Wilson, 2003; Dogan, 2015; Sektnan et al., 2010), longitudinally (e.g., Forget-Dubois et al., 2007; Lansford et al., 2002; Malecki, & Ellio, 2002; McClelland et al., 2000), or by focusing on the associations between predictors and outcome variables during a specific time period (e.g., 50-year time span as shown in Spengler et al.'s study; Masten et al., 2005). Among those studies, limited information was found on what specific behavioral and/or emotional variable impacted an individual's academic performance in school subject areas such as reading and mathematics. Given that previous studies have shown that students' behavioral and emotional characteristics are a leading indicator in their academic performance, examining the long-term relationship between the behavioral/emotional factors and possible change in academic performance can provide a more comprehensive understanding of the roles of students' behaviors and emotions on their academic achievement.

Accordingly, the primary purpose of this study was to examine the longitudinal associations between behavioral and emotional characteristics and student academic performance. With the primary purpose, there were two main questions posed: (a) are the behavioral and emotional characteristics of middle school students associated with their academic achievement? and (b) are the impacts of emotional and behavioral characteristics different according to academic subject? The first question was to investigate possible associations between behavioral and emotional characteristics and student achievement across different grades across times. In order to address the second research question, the associations between behavioral and emotional characteristics and student achievement were examined through two academic areas: Korean language arts and mathematics. Regarding the first question, the current study includes longitudinal data from elementary school-aged to secondary school-students, as they are in a developmentally and academically meaningful stage (i.e., adolescence) and they experience changes in psychological and social domains at secondary school ages (Spear, 2000). In particular, to explore the influence of students' behavioral and emotional characteristics on student achievement over time, a growth modeling approach was applied for the current study, because a growth model allows for tracking the information on the amount of change in student achievement as well as when a critical change has been observed during a given time period (Muthén & Khoo, 1998; Ployhart & Hakel, 1998). Regarding the second question, as previous research has shown that the negative influences of behavioral characteristics have been found in different academic areas (Langberg & Becker, 2012), the current study attempts to examine the growth of student achievement in two major academic subjects.

Methods

Data source

National, longitudinal data

The current study used the data extracted from the public archive of the National Youth Policy Institute (NYPI)'s Korean Students and Youth Panel Survey (KCYPS) (see http://archive.nypi.re.kr/ for more information on data collection procedures and survey items employed). The KCYPS included a set of data from a seven-year longitudinal survey of 7,071 Korean elementary and middle school students. Areas and domains of the survey included the following: personal development (i.e., physical development, intellectual development, social and emotional development, behavioral development); developmental environments (i.e., family environment, peer relationship, community environment, cultural environment); and other demographic information (e.g., gender, living area, parental education, socioeconomic status [SES] level).

Target participants

In order to answer the research questions proposed for the current study, student personal development data (i.e., emotional and behavioral characteristics) and achievement data (i.e., Korean language arts, mathematics) of seventh graders collected in 2013 through 2015 (in ninth grade) was analysed. According to the Korean public education system, the structure of public schools is divided as follows: Six years of elementary school (i.e., first to sixth grade of elementary school), followed by three years of middle school (i.e., seventh to ninth grade) and then three years of high school (i.e., tenth to twelfth grade). The current study included only middle school students (age 12- or 13-years-old for the seventh graders and age 15- or 16-years-old for the ninth graders according to western years), and these grades correspond roughly to grades seven to nine in the US education system.

Data collection procedures

For the KCYPS, the data on participants' emotional and behavioral characteristics were collected using interviews and surveys based on the following procedure: First, trained data collectors and interviewers met with students at a pre-identified location and time and confirmed participant identity; second, the interviewers provided informed consent and instructions for completing the survey and distributed the survey to youths and parents to complete independently; third, interviews were conducted with youths and parents; fourth, survey and interview responses were reviewed and organized using a coding scheme. If there were any noted errors in the surveys, revisions were requested.

Data analysis: Procedures and models

Data extraction and treatment of missing data

As the primary purpose of this study was to identify the associations between emotional/behavioral characteristics (i.e., attention, aggressiveness, behavioral control, social withdrawal, depression, self-esteem) and academic achievement (i.e., Korean language arts and mathematics), the following procedure was conducted for extracting target data. First, according to the research questions and conceptual framework developed by the literature review, six latent factors were selected that might be expected to be found: Attention, aggressiveness, behavior control, social withdrawal, depression, and self-esteem. A total of 36 survey items presenting the selected latent factors and two achievement variables with three time-points (Korean language arts, mathematics) were selected. Second, items of each factor were determined by the selected scales in the KCYPS manuals, as well as a content specialist's opinion and previous studies as follows: (a) the attention factor was comprised of seven questions; (b) the depression factor contained ten questions (see Cho & Lim, 2003; Kim, Kim, & Won, 1983 for validity and reliability information); (c) the aggressiveness factor had six questions; (d) behavioral control included five questions; (e) the social withdrawal factor consisted of five questions; (f) self-esteem consisted of ten questions (see Cho & Lim, 2003; Kim et al., 1983 for psychometric properties of the selected scales and survey items); and (g) the academic achievement factor consisted of Korean language arts and mathematics. Third, response scales were recorded using a four-point Likert-type scale (1: strongly no; 2: relatively no; 3: relatively yes; 4: strongly yes) for all latent traits. For all items, the lower scores students reported, the more positive behavioral characteristics were presented. Responses for Korean language arts (variable name: INT1B01w3) and mathematics (variable name: INT1B02w3) were recorded using an eight-point Likert-type scale ( < 64 score = 1; 65–69 score = 2; 70–74 score = 3; 75–79 score = 4; 80–84 score = 5; 85–89 score = 6; 90–95 score = 7; 95–100 score = 8). Appendix A provides information on items of each scale and Cronbach-alphas of each scale used for the current study. A list-wise deletion for respondents with missing values was applied in this study. In total, 1,874 respondents were used as complete data for this study. The sample size for the growth model met the minimum sample size rule for the SEM study (Bentler & Chou, 1987).

Application of growth model

This study established a growth model to estimate the influences of students' emotional/behavioral characteristics on their reading or mathematics achievement, particularly at middle school entry. The growth model is a specific model of multilevel models, in which repeated observations from an individual are level-1 variables and the latent traits of individuals are the level-2 variables (Bliese & Ployhart, 2002; Chan, 2002). In general, a growth model approach is used to track progress in student achievement and it provides an estimate to determine relative growth over time (Muthén & Khoo, 1998; Ployhart & Hakel, 1998). For the current study, the growth model was used to examine the influence of behavioral and emotional characteristics of the participants on their academic achievement at each time point (i.e., in 2013, 2014, and 2015), as well as how those behavioral and emotional variables impacted the changes across three time-points (i.e., change between 2013–2014; change between 2014–2015).

Three available time-points from KCYPS (2013, 2014, 2015) were extracted to formulate a growth model. This study was particularly designed to predict the intercept (spring of 2013) and slope (spring of seventh grade to spring of ninth grade) terms when students' growth trajectories were estimated. Because the three estimated time points (2013, 2014, 2015) were equally distributed, the value of the first time point was set to 0 and then the subsequent time points were set to 1 and 2, which resulted in the value of 0, 1, and 2 for 2013, 2014, and 2015 test administration, respectively.

The structural equation modeling (SEM) approach was used for estimating growth models. SEM for this study was implemented by a Mplus program (Muthén & Muthén, 1998–2017). The maximum likelihood parameter estimates with robust standard errors (MLR) was applied to estimate parameters for the intercept and slopes. Since multiple responses from an individual will bring some degree of non-independence in the responses (Kenny & Judd, 1986), MLR method that is robust to non-normality and non-independence of observations is appropriate. In order to estimate the model fit, the minimum fit function chi-square statistic ( $χ^{2}$ ), the root mean squared error (RMSE), the comparative fit index (CFI), and the Tucker-Lewis index (TLI), were selected as measures of goodness of fit, as these indices have been used as the most reliable measures of model fit for SEM (MacCallum, Browne & Sugawara, 1996). The cut-off value for the model fit indices applied for the current study was as follows: p value > .05 for $χ^{2}$ ; ≥ .95 for TLI, ≥ .90 for CFI; < 0.08 for RMSE.

The growth model was described as a second order model. The first level is the repeated observations over time and second level is the individual/child level. The first level equation specified the score at each time point and is described as follows:

Y_{ti} = π_{oi} + π_{1 i} t + e_{ti}

(1)

For i = 1,2,3,…, n students, where t is the time and $π_{0 i}$ is the initial status of the child at Time 0, $π_{1 i}$ is the linear slope, indicating the growth rate of person i from initial time point. $e_{ti}$ is the measurement error at the first level and is assumed to be normally distributed with a mean of 0 and variance of 1.

At the second level, the intercept and slope terms of the regression equation become criterion variables and can be predicted by a set of child-level characteristics. The second level equation (for the full independent variables) for the Korean initial status and growth rate parameters are as follows:

\begin{matrix} π_{0 i} = β_{00} + β_{01} (Female) + β_{02} (Att) + β_{03} (Agg) \\ + β_{04} (Soc) + β_{05} (Dep) + β_{06} (Self) + β_{07} (Beh) + r_{0 i} \\ π_{1 i} = β_{10} + β_{11} (Female) + β_{12} (Att) + β_{13} (Agg) \\ + β_{14} (Soc) + β_{15} (Dep) + β_{16} (Self) + β_{17} (Beh) + r_{1 i} \end{matrix}

(2)

where a β estimates the effect of a predictor on a child's initial status and growth rate and

r_{0 i}

and

r_{1 i}

are random error. This model centered the grand-mean for the all predictors so that the intercept parameters (

β_{00}

and

β_{01}

) could be interpreted. For instance, the

β_{00}

can be interpreted as the estimated initial score for the boy in grade 4 who has an average initial attention (Att), aggressiveness (Agg), social withdrawal (Soc), depression (Dep), self-esteem (Self), and behavior control (Beh) scores.

Predictor models

Four models of the effects of intercepts and slopes from spring 2013 in the seventh grade to spring 2015 of the ninth grade of the participants were estimated for Korean language arts and mathematics achievement. This study entered the predictors sequentially. Model 1 included only the students whose status was entered as female. Model 1 investigated the initial and slope difference between males and females over time. Model 2 added the behavioral/emotional variables of attention, aggressive, social withdrawal, and depression. Model 2 evaluated to what extent emotional/psychological variables predict the growth trajectories of students' academic achievement. Model 3 added the self-esteem variable and examined the effect of self-esteem on a participant's initial position and slopes on academic achievements. Model 4 added the behavioral control variable and estimated the effect of behavioral control on a participant's initial position and slopes on academic achievements. Model 4 can be expected to strongly predict a child's later achievement in academics. These models investigated whether (a) attention, aggressiveness, social withdrawal, and/or depression affect academic achievement or (b) students with either high self-esteem or behavioral control grow in academic achievement.

Results

Descriptive statistics

The final sample contained 1,874 students. The descriptive statistics for six predictor variables (i.e., attention, aggressiveness, behavioral control, social withdrawal, depression, self-esteem) are presented in Table 1. The means of scores for each predictor variable ranged from 2.568 (behavioral control) to 3.217 (depression) and the value of variance ranged from 0.138 (self-esteem) to 3.217 (social withdrawal). The descriptive statistics examined whether the data met the assumptions of growth modeling. The assumption of multivariate normal distribution of observations was met because neither the skewedness nor the kurtosis was significant for any variables in this study.

Table 1.

Descriptive statistics for six predictor variables.

Variable Names	Mean	Variance	Median	Skewedness	Kurtosis
Attention	2.822	0.295	2.857	−0.096	0.280
Aggressiveness	3.065	0.293	3.000	−0.145	−0.077
Social withdrawal	2.782	0.560	2.600	−0.044	−0.596
Depression	3.217	0.305	2.800	−0.499	0.285
Self-esteem	2.617	0.138	3.200	0.306	0.245
Behavioral control	2.568	0.286	2.600	0.102	0.117

Note: Response scales were recorded using a four-point Likert-type scale (1: strongly no; 2: relatively no; 3: relatively yes; 4: strongly yes).

Figure 1 shows the students' Korean language arts and mathematics achievement growth trajectories from 2013 through 2015. The mean of Korean language arts scores without putting any predictors increased from 4.45 in 2013 to 4.67 in 2014, and to 5.14 in 2015. The mean of mathematics scores increased from 4.24 in 2013 to 4.29 in 2014, and to 4.54 in 2015. The mean of Korean language arts scores was relatively higher than that of mathematics scores and accelerated in 2015 compared to mathematics scores.

Figure 1.

The change of means scores for Korean language arts and mathematics.

Growth trajectory estimates for Korean language arts and mathematics and the influences of behavioral/emotional variables

Table 2 and 3 presented the estimates of growth trajectory models for students' Korean language arts achievement. The models estimated students' achievement in the Korean language arts in the spring of the seventh grade (intercept) and its growth for three years, i.e., from the seventh grade to the ninth grade (the slope). Model 1 in Table 2 compared the growth of male and female students' achievement. Model 2 added four emotional/behavioral variables (attention, aggressiveness, social withdrawal, depression) to see if these variables predicted the growth of Korean language arts achievement. Model 3 added self-esteem variable to examine if this variable predicted the growth. Model 4 added behavioral control to predict the growth of Korean language arts.

Table 2.

Korean language arts achievement growth models.

	Model 1		Model 2		Model 3		Model 4
	I	S	I	S	I	S	I	S
Intercept	2.618*	0.508*	0.863*	−0.028	0.186	−0.287	−0.335	−0.386*
(Gender) Female	0.004	0.301*	0.526*	0.169*	0.525*	0.169*	0.605*	0.184*
Attention			0.504*	0.150*	0.398*	0.110*	0.017	0.037
Aggressiveness			0.193	−0.051	0.203	−0.047	0.210	−0.045
Social withdrawal			0.054	−0.015	−0.058	−0.058	−0.042	−0.055
Depression			0.190	−0.030	0.112	−0.060	0.029	−0.076
Self-esteem					0.570*	0.218*	0.262*	0.160*
Behavioral control							0.972*	0.184*
Model fit indices	$χ^{2} (20)$ = 313.85* RMSE = 0.089 CFI = 0.884 TLI = 0.861		$χ^{2} (12)$ = 201.39* RMSE = 0.092 CFI = 0.925 TLI = 0.850		$χ^{2} (10)$ = 169.51* RMSE = 0.092 CFI = 0.937 TLI = 0.849		$χ^{2}$ = 20.36* RMSE = 0.029 CFI = 0.995 TLI = 0.985

Note: *p < 0.05.

Table 3.

Mathematics achievement growth models.

	Model 1		Model 2		Model 3		Model 4
	I	S	I	S	I	S	I	S
Intercept	2.960*	0.524*	1.456*	−0.450*	0.751	−0.682*	0.185	−0.815*
(Gender) Female	−0.016	0.048	−0.117	0.172*	−0.117	0.172*	−0.031	0.192*
Attention			0.557*	0.036	0.446*	0.000	0.031	−0.098
Aggressiveness			0.220	−0.018	0.230	−0.014	0.236	−0.013
Social withdrawal			0.080	0.003	−0.037	−0.035	−0.019	−0.031
Depression			0.142	0.091	0.061	0.064	−0.028	0.044
Self-esteem					0.595*	0.195*	0.261	0.116
Behavioral control							1.054*	0.248*
Model fit indices	$χ^{2} (20)$ = 262.69* RMSE = 0.080 CFI = 0.909 TLI = 0.890		$χ^{2} (12)$ = 176.10* RMSE = 0.085 CFI = 0.938 TLI = 0.876		$χ^{2} (10)$ = 154.21* RMSE = 0.088 CFI = 0.946 TLI = 0.870		$χ^{2} (8)$ = 18.52* RMSE = 0.026 CFI = 0.996 TLI = 0.988

Note: *p < 0.05.

Growth trajectory estimates for Korean language arts

In the unconditional model where no predictors were added to the growth equation, the intercept (2.618, p < 0.05) and slope (0.508, p < 0.05) were significant, which was interpreted to mean that growth from seventh grade to ninth grade existed and students' scores in Korean language arts were presented as positive growth over three years (as shown in Figure 1). The correlation between intercept and slopes for the unconditional model was not significant, r = 0.059, which showed that the score of seventh grade achievement was not correlated with their growth from the seventh to ninth grade and that achievement in the seventh grade did not predict significant growth from the seventh to the ninth grade. According to the Model 1 results where the gender variable was added to the growth equation, the growth rate of the male students was lower than the female students, even though their initial achievement was not significantly different, intercept = 0.004 and slope = 0.301, p < 0.05. Model 2 results showed that both intercept and slopes for the attention variable was significant, 0.526, p < 0.05 for intercept and 0.169, p < 0.05 for slope, which indicated that students who have better attention in the seventh grade are likely to show more growth in Korean language arts scores from the seventh grade to ninth grade. In contrast, aggressiveness, social withdrawal, and depression did not predict either initial Korean language arts achievement (intercept = 0.193, 0.054, and 0.190 respectively) or growth over time (slope = -0.051, -0.015, and -0.030 respectively). In Model 3, the female variable and attention still predicted both initial achievement (intercept = 0.525, p < 0.05 for gender; intercept = 0.398, p < 0.05 for attention) and over-time growth (slope = 0.169, p < 0.05 for gender; slope = 0.110, p < 0.05 for attention). Self-esteem also predicted both initial achievement (intercept = 0.570, p < 0.05) and over-time growth (slope = 0.218, p < 0.05). That is, students with higher self-esteem were likely to have higher Korean language arts scores in the seventh grade and showed more growth in achievement from the seventh grade to the ninth grade. Aggressiveness, social withdrawal, and depression variables predicted neither initial Korean achievement nor over-time growth. In Model 4, The behavioral control variable positively predicted both students' initial achievement scores (intercept = 0.972, p < 0.05) and the growth from the seventh grade to the ninth grade (slope = 0.184, p < 0.05). The inclusion of the behavioral control variable did not substantially mediate the positive effect of the self-esteem variable onto the initial score and growth. However, including the behavioral control variable mediated the effect of attention on intercept and slope in the Model.

Model comparison

From the results of modeling fit indices, the Satorra-Bentler scaled chi-square difference test (Satorra & Bentler, 2010) was used to compare the five models (i.e., unconditional model, Model 1, 2, 3, and 4 as presented above) to determine the model that fits the given data well and best explains the associations among the variables. Chi-square difference test results between all Models were shown as significant, $χ^{2} (8)$ = 116.59, p < 0.00 for Model 1 vs. Model 2; $χ^{2} (2)$ = 31.72.59, p < 0.00) for Model 2 vs. Model 3; $χ^{2} (2)$ = 145.30, p < 0.00 for Model 3 vs. Model 4. Thus, it was determined that Model 4 was an improvement over Models 1, 2, and 3 and Model 4 best fits the data as a favorite model.

Growth trajectory estimates for mathematics

Both the intercept (2.960, p < 0.05) and slope (0.524, p < 0.05) in the unconditional model where no predictors were added were significant, which indicated that the growth from seventh to ninth grade was presented in the unconditional model (Figure1). The correlation between intercept and slopes for the unconditional model (i.e., the model using no predictors) was 00.029 and no correlation was found between the achievement level in the seventh grade and the over-time growth rate from the seventh to ninth grade. Model 1 indicated that gender did not have any effect on the intercept (-0.016) and slopes (0.048) in mathematics achievement, indicating that there was no difference between female and male students' mathematics achievement in the seventh grade as well as the over-time growth from the seventh to ninth grade. Model 2 investigated whether the behavioral/emotional variables predicted student growth in mathematics achievement. According to the attention variable's results, the intercept was significant as 0.557, p < 0.05, but the slope was not. This was interpreted to mean that students with higher attention showed higher mathematics achievement in the seventh grade, but this did not predict the over-time growth of the mathematics achievement. The inclusion of behavioral/emotional variables (i.e., attention, aggressiveness, social withdrawal, and depression) in Model 2 allowed female students to have over-time growth in Model 2, slope = 0.172, p < 0.05, indicating that female students' behavioral and emotional variables can affect their growth in mathematics. In Model 3, the attention variable still predicted only mathematics achievement in the seventh grade, intercept = 0.446, p < 0.05, but it did not predict over-time growth. The intercept and slope of the self-esteem variable were both significant, 0.595 and 0.195, p < 0.05 respectively, indicating that students with high self-esteem show higher scores in the seventh grade and these students were very likely to have positive growth from the seventh to ninth grade. Model 4 results showed that both the intercept and slope of the behavioral control were significant, 1.054 and 0.248, p < 0.05 respectively, indicating that behavioral control positively predicted students' mathematics achievement in the seventh grade and their over-time growth. The inclusion of this variable mediated the impacts of the attention and self-esteem variables on both intercept and slope in the Model, with the result that only the behavioral control factor significantly predicted both the intercept and slope of the growth model.

Model comparison

According to the model comparison test, all the comparisons were shown as significant, $χ^{2} (8)$ = 84.45, p < 0.00) for Model 1 vs. Model 2; $χ^{2} (2)$ = 22.15, p < 0.00 for Model 2 vs. Model 3; $χ^{2} (2)$ = 138.41, p < 0.00 for Model 3 vs. Model 4. Thus, Model 4 was proven to be the best fit to the given data and explain the associations among the variables well.

Discussion

The primary purpose of this study was to investigate the associations between middle school students' behavioral and emotional characteristics and the changes in their academic performance measured using Korean language arts and mathematics achievement scores utilizing a growth modeling approach using nationally representative data collected in South Korea. Survey data of middle school students and six behavioral and emotional variables (i.e., attention, aggressiveness, behavioral control, social withdrawal, depression, self-esteem) were used to estimate how those variables were associated with growth in academic performance over time. The major results can be summarized as follows: First, for both academic subjects, students' initial performance and their over-time growth were not correlated; second, the attention variable affected the difference in Korean language arts achievement for seventh grade as well as over-time growth from the seventh to ninth grade, whereas it only affected the difference in the mathematics achievement for the seventh graders; third, for both subjects, the self-esteem variable affected the difference in the seventh grade as well as over-time growth from the seventh to ninth grade; fourth, when the self-esteem variable was added to the model, the influence of the attention variable was lessened for both subjects; fifth, the inclusion of the behavioral control variable reduced the influence of the self-esteem variable for both subjects and students' academic performance in the seventh grade as well as students' over-time growth were impacted by behavioral control. In the following section, these results are further discussed.

The relationship between initial performance and over-time growth

As shown in the unconditional model, when no variables were entered into the growth model, students' academic performance in the seventh grade was not correlated to the growth from the seventh to ninth grade. In general, there has been a prevailing argument that students' initial performance is a strong predictor of their rate of change in academic performance, which is often called as a Matthew Effect in an education setting (Stanovich, 1991, 2000). According to the Mathew Effect, low-achieving or slow learners often show declining progress in their learning, longitudinally resulting in widening gap in achievement rate between slow learners and high-achieving or fast learners (Stanovich, 1986). Thus, we might assume that students who present low initial performance are highly likely to have less over-time growth. Although the over-time growth estimated over three years might not be adequate enough to address individuals' academic growth in the current study, the data from the current study did not support the relationship between initial performance and students' over-time growth. In particular, many studies on the relationship between students' initial performance and their learning trajectories or growth has focused on primary school students (e.g., Burger, 2014; Choi, Elicker, Christ, & Dobbs-Oates, 2016; McClelland, Acock, & Morrison, 2006; Penno, Wilkinson, & Moore, 2002; Shaywitz et al., 1995), whereas this study might provide evidence that runs counter to the Matthew Effect, indicating that middle school students' learning growth can show some variation irrespective of their initial performance.

The influence of behavioral and emotional characteristics

For both Korean language arts and mathematics, the attention variable positively predicted middle school students' academic performance, which is consistent with cumulative, strong evidence on the impact of an individual's attention on his/her academic performance. What this study adds to the existing knowledge is that the attention variable was a predictor of a student's rate of change in academic performance of Korean language arts. That is, the attention variable played an important role in making a difference among the seventh graders as well as predicting students' growth. As revealed in the current study, the influence of attention, particularly for middle school students, has been noted in some literature (e.g., Latimer et al., 2003; Schultz, Evans, & Serpell, 2009) and their findings consistently indicated that a deficit in attention for middle school students resulted in academic decline or even failure in academic performance.

Another variable that showed a strong effect on student academic performance was self-esteem. Of interest, the estimated effect of self-esteem was stronger than the attention variable for both subjects (i.e., Korean language arts, mathematics), and in particular it predicted both the initial performance and the growth for the Korean language arts. Such findings are also supported by the overwhelming body of the studies that emphasize the impact of self-esteem on the academic achievement of middle school students ( Trautwein, Lüdtke, Köller, & Baumert, 2006; Zimmerman & Schunk 2001). For example, in a study conducted by Trautwein et al. (2006), the data of seventh graders from East and West Germany showed a positive relationship between self-esteem and achievement in German and mathematics, suggesting that students' evaluation of themselves can impact on their learning outcomes. Another study that involved Estonian adolescents also showed a strong association between self-esteem and their school performance estimated through Grade Point Average (GPA) (Pullman & Allik, 2008). Along with these findings from the international data, the current study also supported the evidence on the relationship between self-esteem and academic achievement of middle school students. Furthermore, this study showed not only the impact of self-esteem on academic achievement in a cross-sectional way, but also its impact on over-time growth, implying that an individual's attitude towards the self can impact on their ongoing learning outcomes.

An individual's behavioral control refers to his or her perceptions of his or her ability to regulate a situation and/or perform a specific behavior in socially acceptable ways (Chang, Shaw, & Cheong, 2015). Previous research has revealed that age appropriate behavioral control was highly associated with successful interactions with others for school-aged students (Bronson, 2000); there has been relatively little evidence on the relationship between a student's behavioral control and its impact on academic performance. For the current study, the items to represent the behavioral control factor were related to a learner's behavioral control when studying or self-regulation toward task completion (e.g., I study to the end even though studying is boring and uninteresting; I concentrate on my study until I am finished with what I was studying). According to the results, this variable impacted on the seventh graders' achievement as well as their growth from the seventh to ninth grade for both Korean language arts and mathematics. Therefore, according to these results, middle school students' ability to control their learning behaviors can contribute to their growth in learning. Furthermore, in addition to the previous studies which focused on the association of behavioral control with social skills (e.g., interactions with peers; Bronson, 2000 Chang et al. 2015), the findings from the current study also showed that behavioral control can impact on middle school students' academic growth.

Implications for practice

As a way to examine the measurable change in academic achievement from the seventh to the ninth grade according to the behavioral and emotional factors, the findings of the current study can help better understand the possible link between those behavioral variables and academic performance over that specific period of time (Caruana, Roman, Hernández-Sánchez, & Solli, 2015). From the current study, middle school students' aggressiveness, depression, and social withdrawal did not impact on their academic achievement and over-time growth; rather, their attention, self-esteem, and behavioral control showed a high impact on their academic achievement and over-time growth for both Korean language arts and mathematics. That is, it seems fair to say that middle school students' academic achievement was more influenced by their positive intrapersonal characteristics rather than their negative traits over prolonged periods of time. Also, the impact of such positive characteristics was not a one-off thing; the influence reached to their learning growth as well.

These findings can offer some benefit for educating and supporting middle school students, providing information on the behavioral and emotional characteristics as relevant predictors for academic achievement, which can give students, families, and educators a more comprehensive understanding of the relationship between intrapersonal characteristics and achievement. Considering these results, teachers and other related service providers may find justification for and value in promoting self-esteem and behavioral control, understanding that gains in achievement will likely be rendered while simultaneously promoting the confidence and efficacy of students. This might also be reciprocated to impact achievement to a greater degree over time. Furthermore, in the current study there were insignificant relations between negative behavioral and emotional characteristics and middle school students' over-time growth measured through two academic subjects. This should not be interpreted to mean that practitioners and teachers can neglect these negative behavioral and emotional characteristics. As strong evidence from the cumulative studies still support the relationship between those factors and student performance (e.g., Fathi-Ashtiani, Ejei, Khodapanahi, & Tarkhorani, 2007; Park et al., 2018; Rowe, Zimmer-Gembeck, & Hood, 2016; Sun & Shek, 2012), educators need to consider these factors as important indicators.

Limitations and directions for future research

This study, although promising, does present some limitations. First, the KCYPS database, however relevant as a large data base, is somewhat limited in terms of scales and measures related to the selected variables. For the current study, six variables (i.e., attention, aggressiveness, social withdrawal, depression, self-esteem, and behavioral control) were applied as predictors and a few variables had a limited number of items to be constructed as a factor (e.g., only five items for social withdrawal and behavioral control). Although these variables were established based on psychometric properties (i.e., reliability) during the data collection, more items to build a stronger factor can guarantee more accuracy in terms of revealing the behavioral and emotional characteristics. Second, only three intervals from 2013, 2014, and 2015 were used to develop a growth model, which may result in a prediction based on inadequate basis for studying change (Rogosa, Brand, & Zimowski, 1982). Thus, data with more time points would provide a better picture of students' learning growth over time. Third, this study did not consider SES and cognitive ability (e.g., IQ) that previous studies have demonstrated as strong predictors of academic achievement. Accordingly, future studies that consider these variables can add information on how SES and students' cognitive ability impact on and interact with behavioral and emotional characteristics to examine their roles in academic achievement. Finally, as this study used data from South Korea, the findings should be interpreted with caution, particularly when applied to different countries or regions. There may be compounding factors that might have influence on the associations between behavioral and student academic performance related variables; thus, having peer groups from different countries will provide more promising information.

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) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This study was supported through internal grants by the University of Tennessee Knoxville and Hallym University Research Fund (HRF-201808-001).

Author biographies

Yujeong Park, PhD, is an Assistant Professor in the Special Education Program, Department of Theory and Practice in Teacher Education at the University of Tennessee, Knoxville.

Dong Gi Seo, PhD, is an Assistant Professor in the Department of Psychology at Hallym University, Chuncheon-si, Gangwon-do, South Korea.

Jaekook Park, PhD, is a Professor in the Department of Special Education at Pusan National University, Pusan, South Korea.

Byungkeon Kim, PhD, is an Assistant Professor in the Special Education Program at the Graduate School of Education, Soon Chun Hyang University, South Korea.

Jeongwook Choi is a graduate student in the Department of Psychology at Hallym University, Chuncheon-si, Gangwon-do, South Korea.

References

Achenbach

T. M.

(1966) The classification of children's psychiatric symptoms: A factor-analytic study. Psychological Monographs: General and Applied 80(7): 1–37.

Achenbach

T. M.

(1978) The child behavior profile: I. Boys aged 6–11. Journal of Consulting and Clinical Psychology 46(3): 478–488.

Ansary

N. S.

Luthar

S. S.

(2009) Distress and academic achievement among adolescents of affluence: A study of externalizing and internalizing problem behaviors and school performance. Development and Psychopathology 21(1): 319–341. doi:10.1017/S0954579409000182.

Baker

J. A.

Grant

Morlock

(2008) The teacher-student relationship as a developmental context for students with internalizing or externalizing behavior problems. School Psychology Quarterly 23(1): 3–15. doi:10.1037/1045-3830.23.1.3.

Barriga

A. Q.

Doran

J. W.

Newell

S. B.

Morrison

E. M.

Barbetti

Robbins

B. D.

(2002) Relationships between problem behaviors and academic achievement in adolescents: The unique role of attention problems. Journal of Emotional and Behavioral Disorders 10(4): 233–240. doi:10.1177/10634266020100040501.

Bean

R. A.

Bush

K. R.

McKenry

P. C.

Wilson

S. M.

(2003) The impact of parental support, behavioral control, and psychological control on the academic achievement and self-esteem of African and European American adolescents. Journal of Adolescent Research 18(5): 523–541. doi:10.1177/0743558403255070.

Bentler

P. M.

Chou

C. P.

(1987) Practical issues in structural modeling. Sociological Methods and Research 16(1): 78–117.

Bliese

P. D.

Ployhart

R. E.

(2002) Growth modeling using random coefficient models: Model building, testing, and illustrations. Organizational Research Methods 5(4): 362–387. doi:10.1177/109442802237116.

Bronfenbrenner, U., & Morris, P. A. (1998). The ecology of developmental processes. In W. Damon, & R. M. Lerner (Eds.), Handbook of child psychology: Theoretical models of human development (pp. 993–1028). Hoboken, NJ: John Wiley & Sons Inc.

10.

Bronson, M. (2000). Self-regulation in early childhood: Nature and Nurture. New York, NY: Guilford Press.

11.

Burger

(2014) Effective early childhood care and education: Successful approaches and didactic strategies for fostering child development. European Early Childhood Education Journal 23(5): 743–760. doi:10.1080/1350293X.2014.882076.

12.

Capros, J., Cetera, Co., Ogden, L., & Rossett, K. (2002). Improving student's social skills and achievement through cooperative learning. ERIC Document Reproduction Service No. ED468873.

13.

Caruana

E. J.

Roman

Hernández-Sánchez

Solli

(2015) Longitudinal studies. Journal of Thoracic Disease 7: 537–540. doi:10.3978/j.issn.2072-1439.2015.10.63.

14.

Chan, D. (2002). Latent growth modeling. In F. Drasgow, & N. Schmitt (Eds.), Measuring and analyzing behavior in organizations: Advances in measurement and data analysis (pp. 302–349). San Francisco, CA: Jossey-Bass.

15.

Chang

Shaw

D. S.

Cheoung

(2015) The development of emotional and behavioral control in early childhood: Heterotypic continuity and relations to early school adjustment. Journal of Child and Adolescent Behavior 3(3): 204–213. doi:10.4172/2375–4494.1000204. .

16.

Cho

Lim

(2003) Development and validation of emotional or behavioral problems scale. Korean Journal of Counseling and Psychotherapy 15(4): 729–746.

17.

Choi

J. Y.

Elicker

Christ

S. L.

Dobbs-Oates

(2016) Predicting growth trajectories in academic learning: Evidence from growth curve modeling with head start children. Early Childhood Research Quarterly 36: 244–258. doi:10.1016/j.ecresq.2015.12.017.

18.

Chu

S.-Y.

Baker

(2015) The effects of video self-modeling on high school 159 students with emotional and behavioral disturbances. Preventing School Failure 59: 207–216. doi:10.1080/1045988X.2014.903465.

19.

Coie, J. D. (1996). Prevention of violence and antisocial behavior. In R. Peters, & R. McMahon (Eds.), Preventing childhood disorders, substance abuse, and delinquency (pp. 1–18). Thousand Oaks, CA: Sage.

20.

DiPrete

T. A.

Jennings

J. L.

(2012) Social and behavioral skills and the gender gap in early educational achievement. Social Science Research 41(1): 1–15. doi:10.1016/j.ssresearch.2011.09.001.

21.

Dogan

(2015) Student engagement, academic self-efficacy, and academic motivation as predictors of academic performance. Anthropologist 20(3): 553–561. doi:10.1080/09720073.2015.11891759.

22.

Fathi-Ashtiani

Ejei

Khodapanahi

M. K.

Tarhorani

(2007) Relationship between self-concept, self-esteem, anxiety, depression and academic achievement in adolescent. Journal of Applied Sciences 7(7): 995–1000. doi:10.3923/jas.2007.995.1000.

23.

Forget-Dubois

Lemelin

J. P.

Boivin

Dionne

Séguin

J. R.

Vitaro

Tremblay

R. E.

(2007) Predicting early school achievement with the EDI: A longitudinal population-based study. Early Education and Development 18(3): 405–426. doi:10.1080/10409280701610796.

24.

Forness

S. R.

Kim

Walker

H. M.

(2012) Prevalence of students with EBD: Impact on general education. Beyond Behavior 21(2): 3–11.

25.

Frazier

T. W.

Youngstrom

E. A.

Naugle

R. I.

(2007) The latent structure of attention-deficit/hyperactivity disorder in a clinic-referred sample. Neuropsychology 21: 45–64. doi:10.1037/0894-4105.21.1.45.

26.

Gage

N. A.

Josephs

N. L.

Lunde

(2012) Girls with emotional disturbance and a history of arrest: Characteristics and school-based predictors of arrest. Education and Treatment of Children 35(4): 603–622. doi:10.1353/etc.2012.0026.

27.

Hagan-Burke

Kwok

O. M.

Zou

Johnson

Simmons

Coyne

M. D.

(2011) An examination of problem behaviors and reading outcomes in kindergarten students. The Journal of Special Education 45(3): 131–148. doi:10.1177/0022466909359425.

28.

Kauffman

J. M.

Mock

D. R.

Simpson

R. L.

(2007) Problems related to underservice of students with emotional or behavioral disorders. Behavioral Disorders 33(1): 43–57.

29.

Kellam

S. G.

Werthamer-Larsson

Dolan

L. J.

Brown

C. C. H.

Mayer

L. S.

Rebok

G. W.

, et al.(1991) Developmental epidemiologically based preventive trials: Baseline modeling of early target behaviors and depressive symptoms. American Journal of Community Psychology 19: 563–584.

30.

Keller, J. M. (1983). Motivational design of instruction. In C. M. Reigeluth (Eds.). Instructional-design theories and models: An overview of their current status (pp. 386–434). Hillsdale, NJ: Lawrence Erlbaum Associates.

31.

Kenny

D. A.

Judd

C. M.

(1986) Consequences of violating the independence assumption in analysis of variance. Psychological Bulletin 99(3): 422–431.

32.

Kim, K., Kim, J., & Won, H. (1983). Korean Manual of Symptom Checklist-90-Revision. Seoul: Choong Ang Aptitude Publishing Co.

33.

Langberg

J. M.

Becker

S. P.

(2012) Does long-term medication use improve the academic outcomes of youth with attention-deficit/hyperactivity disorder? Clinical Child and Family Psychology Review 15(3): 215–233. doi:10.1007/s10567-012-0117-8.

34.

Lansford

J. E.

Dodge

K. A.

Pettit

G. S.

Bates

J. E.

Crozier

Kaplow

(2002) A 12-year prospective study of the long-term effects of early child physical maltreatment on psychological, behavioral, and academic problems in adolescence. Archives of Pediatrics and Adolescent Medicine 156(8): 824–830. doi:10.1001/archpedi.156.8.824.

35.

Latimer

W. W.

August

G. J.

Newcomb

M. D.

Realmuto

G. M.

Hektner

J. M.

Mathy

R. M.

(2003) Child and familial pathways to academic achievement and behavioral adjustment: A prospective six-year study of children with and without ADHD. Journal of Attention Disorders 7(2): 101–116. doi:10.1177/108705470300700204.

36.

Masten

A. S.

Roisman

G. I.

Long

J. D.

Burt

K. B.

Obradović

Riley

J. R.

, et al.(2005) Developmental cascades: Linking academic achievement and externalizing and internalizing symptoms over 20 years. Developmental Psychology 41(5): 733–746. doi:10.1037/0012-1649.41.5.733.

37.

MacCallum

R. C.

Browne

M. W.

Sugawara

H. M.

(1996) Power analysis and determination of sample size for covariance structure modeling. Psychological Methods 1(2): 130–149.

38.

Malecki

C. K.

Elliott

S. N.

(2002) Children's social behaviors as predictors of academic achievement: A longitudinal analysis. School Psychology Quarterly 17(1): 1–23. doi:10.1521/scpq.17.1.1.19902.

39.

McClelland

M. M.

Acock

A. C.

Morrison

F. J.

(2006) The impact of kindergarten learning-related skills on academic trajectories at the end of elementary school. Early Childhood Research Quarterly 21: 471–490. doi:10.1016/j.ecresq.2006.09.003.

40.

McClelland

M. M.

Morrison

F. J.

Holmes

D. L.

(2000) Children at risk for early academic problems: The role of learning-related social skills. Early Childhood Research Quarterly 15(3): 307–329. doi:10.1016/S0885-2006(00)00069-7.

41.

Merrell

Tymms

P. B.

(2001) Inattention, hyperactivity and impulsiveness: Their impact on academic achievement and progress. British Journal of Educational Psychology 71(1): 43–56. doi:10.1348/000709901158389.

42.

Muthén

B. O.

Khoo

S. T.

(1998) Longitudinal studies of achievement growth using latent variable modeling. Learning and individual differences 10(2): 73–101.

43.

Muthén, L. K., & Muthén, B. (1998-2017). Mplus User's Guide. Eighth Edition. Los Angeles, CA: Muthen & Muthen.

44.

Park

Seo

D. G.

Moore

E. J.

Kim

(2018) What contributes to low achievement of middle school students: Evidence from multigroup structural equation modeling. The Journal of Educational Research 111(4): 404–416. doi: 10.1080/00220671.2017.1284040.

45.

Penno

J. F.

Wilkinson

I. A.

Moore

D. W.

(2002) Vocabulary acquisition from teacher explanation and repeated listening to stories: Do they overcome the Matthew effect? Journal of Educational Psychology 94(1): 23–33. doi:10.1037/0022-0663.94.1.23.

46.

Pekrun

(2006) The control-value theory of achievement emotions: Assumptions, corollaries, and implications for educational research and practice. Educational Psychology Review 18(4): 315–341. doi:10.1007/s10648-006-9029-9.

47.

Pierson

M. R.

Carter

E. W.

Lane

K. L.

Glaeser

B. C.

(2008) Factors influencing the self-determination of transition-age youth with high-incidence disabilities. Career Development for Exceptional Individuals 31(2): 115–125. doi:10.1177/0885728808317659.

48.

Ployhart

R. E.

Hakel

M. D.

(1998) The substantive nature of performance variability: Predicting interindividual differences in intraindividual performance. Personnel Psychology 51: 859–901. doi:10.1111/j.1744-6570.1998.tb00744.x.

49.

Pullmann

Allik

(2008) Relations of academic and general self-esteem to school achievement. Personal Individual Differences 45(6): 559–564. doi:10.1016/j.paid.2008.06.017.

50.

Rogosa

D. R.

Brand

Zimowski

(1982) A growth curve approach to the measurement of change. Psychological Bulletin 90: 726–748. doi:10.1037/0033-2909.92.3.726.

51.

Rowe

S. L.

Zimmer-Gembeck

M. J.

Hood

(2016) From the child to the neighborhood: Longitudinal ecological correlates of young adolescents' emotional, social, conduct, and academic difficulties. Journal of Adolescence 49: 218–231. doi:10.1016/j.adolescence.2016.04.001.

52.

Satorra

Bentler

P. M.

(2010) Ensuring positiveness of the scaled difference chi-square test statistic. Psychometrika 75(2): 243–248. doi:10.1007/S11336-009-9135-Y.

53.

Schultz

B. K.

Evans

S. W.

Serpell

Z. N.

(2009) Preventing failure among middle school students with attention deficit hyperactivity disorder: A survival analysis. School Psychology Review 38(1): 14–27.

54.

Sektnan

McClelland

M. M.

Acock

Morrison

F. J.

(2010) Relations between early family risk, children's behavioral regulation, and academic achievement. Early Childhood Research Quarterly 25(4): 464–479. doi:10.1016/j.ecresq.2010.02.005.

55.

Shaywitzm

B. A.

Holford

T. R.

Holahan

J. M.

Fletcher

J. M.

Stuebing

K. K.

Francis

D. J.

Shaywitz

S. E.

(1995) A Matthew effect for IQ but not for reading: Results from a longitudinal study. Reading Research Quarterly 30(4): 894–906. doi:10.2307/748203.

56.

Skinner

E. A.

Wellborn

J. G.

Connell

J. P.

(1990) What is takes to do well in school and whether I've got it: A process model of perceived control and children's engagement and achievement in school. Journal of Educational Psychology 82(1): 22–32. doi:10.1037/0022-0663.82.1.22.

57.

Spear

L. P.

(2000) The adolescent brain and age-related behavioral manifestations. Neuroscience and Biobehavioral Reviews 24: 417–463. doi:10.1016/S0149-7634(00)00014-2.

58.

Spengler

Damian

R. I.

Roberts

B. W.

(2018) How you behave in school predicts life success above and beyond family background, broad traits, and cognitive ability. Journal of Personality and Social Psychology 114(4): 620–636. doi:10.1037/pspp0000185.

59.

Spira

E. G.

Fischel

J. E.

(2005) The impact of preschool inattention, hyperactivity, and impulsivity on social and academic development: A review. Journal of Child Psychology and Psychiatry 46(7): 755–773. doi:10.1111/j.1469-7610.2005.01466.x.

60.

Stanovich

K. E.

(1986) Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quarterly 21: 360–406.

61.

Stanovich

K. E.

(1991) Discrepancy definitions of reading disability: Has intelligence led us astray? Reading Research Quarterly 26: 7–29. doi:10.2307/747729.

62.

Stanovich, K. E. (2000). Progress in understanding reading: Scientific foundations and new frontiers. New York, NY: Guilford Press.

63.

Stuhlman

M. W.

Pianta

R. C.

(2009) Profiles of educational quality in first grade. The Elementary School Journal 109(4): 323–342.

64.

Sun

R. C. F.

Shek

D. T. L.

(2012) Student classroom misbehavior: An exploratory study based on teachers' perceptions. The Scientific World Journal 2012: 1–8. doi:10.1100/2012/208907.

65.

Thompson

(2009) Disruptive behaviours in Barbadian classrooms: Implications for universal secondary education in the Caribbean. Journal of Eastern Caribbean Studies 34(3): 39–58.

66.

Trautwein

Lüdtke

Köller

Baumert

(2006) Self-esteem, academic self-concept, and achievement: How the learning environment moderates the dynamics of self-concept. Journal of Personality and Social Psychology 90(2): 334–349. doi:10.1037/0022-3514.90.2.334.

67.

Trzesniewski

K. H.

Moffitt

T. E.

Caspi

Taylor

Maughan

(2006) Revisiting the association between reading achievement and antisocial behavior: New evidence of an environmental explanation from a twin study. Child Development 77(1): 72–88. doi:10.1111/j.1467-8624.2006.00857.x.

68.

van Lier

P. A.

Vitaro

Barker

E. D.

Brendgen

Tremblay

R. E.

Boivin

(2012) Peer victimization, poor academic achievement, and the link between childhood externalizing and internalizing problems. Child Development 83(5): 1775–1788. doi:10.1111/j.1467-8624.2012.01802.x.

69.

Zimmerman

B. J.

(1989) A social cognitive view of self-regulated academic learning. Journal of Educational Psychology 81(3): 329–339. doi:10.1037/0022-0663.81.3.329.

70.

Zimmerman, B. J., & Schunk, D. H. (2001). Self-regulated learning and academic achievement: Theoretical perspectives. Mahwah, NJ: Lawrence Erlbaum Associates.