Does Rude or Kind Behavior Predict Later Academic Achievement? Evidence From Two Samples of Adolescents

Abstract

This study examined whether adolescent students’ externalizing behavior and prosocial behavior affects their academic achievement (i.e., school grades and standardized test scores) in two major academic domains (i.e., mathematics and German) over and above well-established determinants of school achievement (i.e., prior achievement, intelligence, and domain-specific self-concepts). The study draws on longitudinal data from two independent samples of students (A/B) who were each followed from Grade 7 to Grade 9 (N_A/B = 1143/1345). In each academic domain, structural equation models showed that externalizing behavior negatively predicted and prosocial behavior positively predicted academic achievement. When both behaviors were included simultaneously, prosocial behavior particularly predicted school grades in both domains, whereas externalizing behavior particularly predicted test scores in mathematics. Further analyses did not suggest differences between boys and girls in the effects of social behavior on academic achievement. Implications for future research and educational practice are discussed.

Keywords

externalizing behaviors prosocial behavior problem/risky/antisocial behaviors academic achievement

Externalizing can be defined as aggressive and rule-breaking behavior, such as constant lying, verbal or physical aggression, stealing, disrupting, truancy, or swearing (e.g., Achenbach et al., 2016; Hinshaw, 1992; Stanger et al., 1997). Prior research has indicated that externalizing behavior is negatively related to achievement (e.g., van der Ende et al., 2016; Wentzel, 1993; Zhang et al., 2019). The assumption that initial problem behavior causes subsequent academic difficulties is also known as the adjustment erosion hypothesis (Moilanen et al., 2010) and is a central focus of this research.

If externalizing behavior (as a problematic behavior) is detrimental to achievement, prosocial behavior (as a generally functional behavior) might be positively connected to achievement. Prosocial behavior is characterized as being voluntary oriented toward supporting others’ well-being and includes actions such as helping, sharing, and cooperation (Batson & Powell, 2003; Eisenberg & Miller, 1987; Padilla-Walker & Carlo, 2014). Indeed, prior research has suggested a positive association between prosocial behavior and achievement (e.g., Caprara et al., 2000; Curlee et al., 2019; Wentzel, 1993). The assumption that prosocial behavior can help facilitate learning processes and enhance achievement, which we would refer to as the adjustment accumulation hypothesis, is another central focus of this research.

Although prior studies indicate a negative relation between externalizing behavior and academic achievement and a positive relation between prosocial behavior and academic achievement, it remains unknown whether externalizing and prosocial behavior indeed influence academic achievement when other factors that are known to influence school achievement (i.e., prior achievement, intelligence, and domain-specific self-concepts) are controlled for. Therefore, the central aim of this study was to address this question. In addition, we investigated whether the relations differ between (a) achievement as measured with teacher-assigned school grades and standardized test scores, (b) the mathematical and verbal domains, and (c) boys and girls. We have also investigated whether (d) externalizing and prosocial behavior have unique relations with school achievement when considered simultaneously.

Social Behavior as a Factor That Influences Academic Achievement

There are several explanations for why externalizing and prosocial behavior could affect academic achievement, which are not mutually exclusive. For example, there is empirical evidence that externalizing behavior is detrimental, whereas prosocial behavior is beneficial for the quality of relationships with classmates and teachers (e.g., Nurmi, 2012; van Lier & Koot, 2010; Vannatta et al., 2009). The quality of these relationships has in turn been shown to be important for students’ engagement, and achievement (e.g., Longobardi et al., 2016; Roorda et al., 2017; Weyns et al., 2018).

It is also conceivable that part of the variance shared between social behavior and achievement can be attributed to biases in how school achievement is judged. For example, researchers have discussed whether teachers might be rewarding or penalizing desirable or undesirable classroom behavior by giving grades that, at least to some extent, reflect students’ social behavior (Loveland et al., 2007; Wentzel, 2005). Furthermore, teachers’ grading practices might be unconsciously biased by a halo effect (Thorndike, 1920) that results from teachers’ perceptions of students’ social behavior. Whether intentional or not, there is at least some evidence that grading is negatively affected by externalizing behavior, even when actual achievement is experimentally controlled for (e.g., Krämer & Zimmermann, 2020).

Externalizing Behavior and Academic Achievement

Many studies have suggested that externalizing behavior contributes negatively to academic achievement. As early as 1992, in his widely acclaimed review, Hinshaw concluded that externalizing behavior and underachievement are clearly associated. However, a recently published review questioned the predictive relation between externalizing behavior and academic achievement due to inconclusive empirical evidence (Kulkarni et al., 2021). Indeed, many studies have suggested a negative relation between externalizing behavior and academic achievement (e.g., Deighton et al., 2018; Finn et al., 2008; Zimmermann et al., 2013). However, other studies have shown nonsignificant or inconclusive results (e.g., Caprara et al., 2000; Halonen et al., 2006; Romano et al., 2010). Factors that contribute to the uncertainty surrounding the relation likely consist of the facts that many studies (a) did not directly focus on externalizing behavior but focused instead on related variables, (b) did not control for various other known determinants of school achievement, and (c) investigated either grades or standardized test scores.

Studies that did not directly focus on externalizing behavior but focused on related variables can give only a hint about the relation between externalizing behavior and school achievement. For instance, many studies focused on aggressive behavior (e.g., Caprara et al., 2000; Doctoroff et al., 2006; Loveland et al., 2007), and although aggressive behavior is an externalizing behavior, it does not capture the broader construct of externalizing behavior, which additionally includes rule-breaking behavior (e.g., Achenbach et al., 2016).

Further, only a few studies have investigated the relation between externalizing behavior and academic achievement while controlling for various other factors that have been robustly shown to contribute to academic achievement (e.g., Romano et al., 2010; Wentzel, 1993). But if studies do not control for various other known determinants of school achievement (e.g., students’ intelligence or prior achievement), the extent of a unique contribution of externalizing behavior to academic achievement will remain quite uncertain.

Finally, most previous studies investigated either grades or standardized test scores as the dependent variable (e.g., Doctoroff et al., 2006; Loveland et al., 2007; Miles & Stipek, 2006). Test scores provide a rather objective measure of performance, whereas report card grades capture a mixture of assessments of student performance and other measures (e.g., effort, discipline, and participation in class). Thus, report card grades are more multidimensional measures than standardized test scores are. Due to this difference, it is plausible that externalizing behavior may have different associations with grades and test scores as achievement measures. Although both outcomes are relevant for their own reasons, studies that measure both of them in the same sample are particularly revealing. A few studies have already included both outcomes and have indeed suggested that the effects on school grades exceed those on standardized achievement test scores (Okano et al., 2020; Wentzel, 1993; Zimmermann et al., 2013). These different relations could contribute to the impression of inconclusive results regarding the relation between externalizing behavior and academic achievement and should be investigated further.

Prosocial Behavior and Academic Achievement

The relation between prosocial behavior and academic achievement has been investigated less frequently than the relation between externalizing behavior and academic achievement. Nevertheless, there is evidence that prosocial behavior contributes positively to academic achievement (e.g., Gerbino et al., 2018; Wentzel, 1993; Wentzel & Caldwell, 1997).

Fortunately, many of these studies have investigated the relation between prosocial behavior and academic achievement with prospective designs. For instance, Caprara et al. (2000) found that prosocial behavior in third grade had a positive effect on averaged school grades in eighth grade. Similarly, Gerbino et al. (2018) found a positive effect of prosocial behavior in eighth grade on school grades 5 years later. Covering a similarly long time and addressing not only teacher-assigned grades but also test scores, Curlee et al. (2019) found that prosocial behavior predicted both achievement measures. Studies by DeVries et al. (2018) and Wentzel (1993) also examined both achievement measures, with results pointing to a particular relevance of prosocial behavior for school grades. As for externalizing behavior, the effects of prosocial behavior on grades seem to exceed the effects on test scores as well.

In Wentzel’s (1993) study, prosocial behavior was related to academic achievement even though other strong predictors of school achievement were controlled. For instance, prosocial behavior had unique effects on school grades over and above the significant effects of externalizing behavior, intelligence, academic behavior, family structure, and days absent from school.

In sum, prior studies have lend some support that prosocial behavior contributes to academic achievement. Nevertheless, there is still a need for research on the effect of prosocial behavior on academic achievement due to a lack of studies that use a prospective design and control for the effects of known determinants of school achievement, and moreover, address both school grades and standardized test scores as achievement measures.

The Present Study

We conducted the present study to examine whether adolescent students’ social behavior (i.e., externalizing and prosocial behavior) contributes to academic achievement as assessed with school grades and standardized achievement tests in the mathematical and verbal domains. We investigated these relations separately and simultaneously for the two forms of behavior. In all analyses, we controlled for the influence of well-established determinants of school achievement (i.e., prior achievement, general cognitive abilities, and domain-specific self-concepts).

We decided to include both externalizing and prosocial behavior in our study. Externalizing and prosocial behavior are negatively related but do not represent opposite poles of a single construct. Scholars have suggested that their relation is much more complex (e.g., Kokko et al., 2006; Padilla-Walker et al., 2018; Schröder et al., 2016) and, although the size of the negative correlation has varied, it tends to be rather moderate (e.g., Muris et al., 2003; Vinnick & Erickson, 1992; Wentzel, 1993). Moreover, prosocial behavior is more than the mere absence of externalizing behavior and vice versa (e.g., not behaving aggressively does not allow for conclusions about prosocial behavior; e.g., Kokko et al., 2006).

With externalizing and prosocial behavior representing different but related constructs, it is especially interesting to include them both in an investigation and get closer to their unique effects. Regarding the question of determinants of school achievement, very few studies have investigated the two behaviors simultaneously. Nevertheless, initial studies have hinted at the particular importance of prosocial behavior when externalizing behavior was also taken into account (e.g., Caprara et al., 2000; Montroy et al., 2014; Wentzel & Caldwell, 1997).

Various studies have investigated factors that contribute to students’ academic achievement (see, e.g., Hattie, 2008). A persistent finding was that prior achievement determined subsequent achievement (e.g., Chen et al., 1996; Hattie, 2008). This finding is also known as the Matthew effect. The more quickly expanding knowledge base of those who have more prior knowledge can be explained by facilitated learning, as prior knowledge enables better elaboration of new information (Stanovich, 1986). We controlled for prior achievement in our study to take into account prior existing differences in students’ achievement and, therefore, to get closer to the unique effects of social behavior on academic achievement. Furthermore, the relevance of general cognitive abilities (i.e., intelligence) for academic achievement has often been investigated and has been found to make a substantial contribution (e.g., Roth et al., 2015; Spinath et al., 2006; Spinath et al., 2010). Besides prior achievement and general cognitive abilities, we controlled for students’ domain-specific self-concepts as motivational antecedents of achievement. For example, in Eccles, Wigfield, and colleagues’ expectancy-value theory of achievement motivation, ability self-concepts contribute to the expectation of success, which influences achievement-related choices, effort, persistence, and achievement (e.g., Wigfield & Eccles, 2000). Prior research has shown that self-concepts are positively related to academic achievement (e.g., Hattie, 2008; Huang, 2011; Spinath et al., 2006).

This study was conducted with participants in early adolescence. This time is particularly interesting to examine influences on achievement, as many students are facing the decision to continue or end their general school education. Achievement at this time is, therefore, very important. In addition, adolescence is particularly interesting for our research question because some externalizing behaviors seem to be especially prevalent (e.g., Burt, 2012; Van Lier et al., 2009).

Prior studies indicated that boys tend to show more externalizing behavior than girls, whereas girls tend to show more prosocial behavior than boys (e.g., Adams et al., 1999; Veenstra et al., 2008). Such differences are in line with traditional gender role expectations: externalizing behavior, as a sign of laddishness and assertiveness, corresponding to the traditionally male role, and prosocial behavior, as a sign of caring and nurturing, corresponding to the traditionally female role (e.g., Eagly et al., 2000). The respective gender-atypical behavior might be particularly salient because it contradicts traditional gender role expectations. Depending on gender role typicality of the behavior, gender could thus influence the relation between behavior and achievement. However, previous empirical findings are inconsistent. For example, Loveland et al. (2007) found stronger relations between aggressive behavior and school achievement for female adolescents, whereas Doctoroff et al.’s (2006) results stressed the relevance of both aggressive and prosocial behavior for school achievement among male children. Some studies did not indicate gender differences in these relations (e.g., Deighton et al., 2018; Guo et al., 2018; van der Ende et al., 2016). Thus, we explored whether students’ gender moderates the effects without specific expectations.

Method

Sample and Procedure

The participants in the present study took part in a larger longitudinal multicohort study conducted in Germany, with data collections between 2004 and 2008 (EIKA, 2009). The study was conducted in accordance with APA ethical principles concerning research (American Psychological Association, 2017).

In the present study, we used two cohorts of students (i.e., Samples A and B), who were followed from Grade 7 to Grade 9. We conducted all analyses in both samples separately. This approach has the great advantage that we can take into consideration whether result patterns can be replicated.

The two samples consisted of N_A = 1143 and N_B = 1345 students from 10 schools in socially disadvantaged urban areas covering the whole range of types of secondary schools in Germany’s tracked school system. About a quarter of the students (Sample A = 27.6% and Sample B = 24.6%) belonged to the academic track. Other students belonged to typical vocational tracks (Sample A = 43.4 and B = 75.4), and due to changes in the school system, some students in Sample A belonged to a mixed track (Sample A = 29.0). Students were on average 12 years old when in Grade 7 (M_A = 12.7, SD_A = 0.7; M_B = 12.8, SD_B = 0.8) and correspondingly 2 years older in Grade 9. About half of the students were girls (Sample A = 49.9% and Sample B = 46.8%), and the proportion of students with a migration background (defined by whether at least one parent or the student was not born in Germany) was rather high (Sample A = 51.9% and Sample B = 51.5%). Socioeconomic status as measured with the HISEI (Highest International Socio-Economic Index of Occupational Status; Ganzeboom & Treiman, 1996) on a scale ranging from 16 to 90 was rather low (M_A = 39.2, SD_A = 14.1; M_B = 39.8, SD_B = 14.1; see, e.g., Klieme et al., 2010).

Data were collected for the predictors when the students were in Grade 7. At the beginning of the school year, standardized achievement tests in mathematics and German and general cognitive abilities tests were administered to the students, as well as a questionnaire that captured, for example, demographic data and students’ domain-specific ability self-concepts in mathematics and German. Data collection was conducted by teachers using fully standardized instructions. About 2 months after the school year began, the teacher with the most hours spent teaching the class filled out the questionnaires about externalizing and prosocial behavior for each student in the class. The students’ report card grades in mathematics and German from the end of the previous school year (i.e., Grade 6) were taken from the schools’ registries.

Two years later, at the beginning of Grade 9, the achievement variables were measured again to represent the outcome variables. Achievement tests were administered and the report card grades in mathematics and German from the end of Grade 8 were taken from the schools’ registries.

Participation in the performance tests was obligatory for the students, but filling in the questionnaires was voluntary without compensation.

On average, 17.2% of the values of the model variables in Sample A and 13.0% in Sample B were missing (see Table 1 for the proportion of missing data per variable). Reasons for missing data can be manifold. For example, students may be sick on a day of data collection, may not want to fill in the questionnaires, or change schools between waves of data collection. In both samples of our study, achievement data at the second measurement were relatively often missing. We compared whether students with complete achievement data at the second measurement differed from those with missings on achievement data at the second measurement regarding different background variables. In Sample A, students with complete and students with missings on achievement data at the second measurement had a quite similar proportion of girls (with complete data 48.6% and with missing data 51.4%), students with a migration background (45.4% and 46.3%), and a similar socioeconomic background as measured with the HISEI (Ganzeboom & Treiman, 1996; M = 38.8 and 39.5; SD = 14.0 and 14.5). However, in Sample B, students with missings on achievement variables were less often female (with complete data 48.6% and with missing data 42.6%), had more often a migration background (37.8% and 53.4%), and had a slightly lower HISEI (Ganzeboom & Treiman, 1996; M = 40.5 and 37.5; SD = 14.2 and 13.4).

Table 1.

Descriptive Statistics.

Variable	M (SD) Sample A	M (SD) Sample B	Range Sample A	Range Sample B	Missing data Sample A (%)	Missing data Sample B (%)
Externalizing behavior	0.2 (0.4)	0.2 (0.3)	0.0–1.9	0.0–1.8	11.0	8.8
Prosocial behavior	3.5 (0.9)	3.3 (0.8)	1.0–5.0	1.0–5.0	11.8	10.0
Grade math.	3.7 (1.0)	3.7 (1.0)	1–6	1–6	37.5	20.0
Grade German	3.8 (0.8)	3.8 (0.8)	2–6	1–6	37.6	21.3
Test score math.	2.1 (1.0)	2.0 (1.1)	−0.9–4.8	−0.6–4.9	31.6	24.2
Test score German	2.0 (1.0)	2.0 (0.9)	−0.9–4.8	−0.3–4.3	30.4	25.2
Prior grade math.	3.8 (1.0)	3.6 (1.1)	1–6	1–6	7.2	3.3
Prior grade German	3.8 (0.9)	3.7 (1.0)	1–6	1–6	7.2	3.4
Prior test score math.	1.0 (0.9)	1.2 (1.0)	−2.3–5.0	−2.3–5.0	7.3	4.8
Prior test score German	1.2 (1.1)	1.0 (1.0)	−1.3–5.6	−1.6–5.3	11.4	5.6
General cog. abilities	30.4 (5.4)	30.0 (5.9)	11–44	7–45	8.2	4.1
Self-concept math.	2.8 (0.9)	2.8 (0.9)	1–4	1–4	16.5	23.9
Self-concept German	2.9 (0.8)	2.8 (0.8)	1–4	1–4	18.9	26.0
Gender	49.9% female	46.8% female	—	—	4.3	1.3

Note. Means (M), standard deviations (SD), ranges of the scales, proportion of female students and missing data. Math. = mathematics; cog. = cognitive.

Measures

School grades

Teachers of the respective subjects assigned school grades in mathematics and German. In Germany, the grading system ranges from 1 (outstanding) to 6 (fail). To ease the interpretation of the results, school grades were recoded so that higher scores represented better grades.

Test scores

Items from standardized tests developed within the framework of large-scale educational studies (Lehmann et al., 1999; Lehmann et al., 2002; OECD, 2003; Orpwood & Garden, 1998) were used to measure test scores in the mathematical and verbal domains. We used an anchor-item design to capture school achievement across grades in a domain on the same metric. For the verbal domain, the test focused on reading comprehension. Students had to read several texts and answer questions about them. For mathematics, the tasks corresponded to the curriculum for the students’ grade level and included, for example, fractions or probabilities. For both tests, the answers were dichotomized, and weighted likelihood estimates (WLEs) were computed with the ConQuest software (Wu et al., 1998). The WLE reliabilities were sufficient for both tests (≥.75 in both samples and measurement points).

Externalizing behavior

Measures of students’ social behavior were filled out by the teacher with the most hours in a class. To measure students’ externalizing behavior, the teachers rated 26 items from two subscales from the German version of the Teacher’s Report Form (Achenbach, 1991; Arbeitsgruppe Deutsche Child Behavior Checklist, 1993). The items stemmed from the aggressive behavior (e.g., “attacks people”) and delinquent behavior (e.g., “steals”) subscales. Teachers rated each individual student on the 26 misbehaviors on a 3-point scale by indicating whether a characteristic was not true (0), somewhat or sometimes true (1), or very or often true (2) at the present time or in the recent past. The reliabilities were excellent (Cronbach’s α_A = .96; Cronbach’s α_B = .95). The items were used to build six parcels, consisting of four to five items each, that were used to build the latent variable externalizing behavior. To analyze the measurement model, we conducted confirmatory factor analyses in both samples. The goodness-of-fit indices indicated that both models fit the data well, χ²_A/B (9/9) = 80.5/85.0, CFI_A/B = .98/.98, TLI_A/B = .96/.97, RMSEA_A/B = .09/.08, and SRMR_A/B = .02/.02. The standardized factor loadings were at least .87 and .86 for Samples A and B, respectively.

Prosocial behavior

We used a prosocial behavior scale developed for the current study. Teachers rated their students on six behaviors addressing prosocial behavior in school, such as “is helpful to classmates” or “tries frequently to settle disputes” on a 5-point scale ranging from 1 (not true at all) to 5 (exactly true). The scale thus covers typical behaviors that were also used in other instruments to measure prosocial behavior (e.g., Goodman, 1997; Tremblay et al., 1992). The reliabilities of the scale were excellent in both samples (Cronbach’s α = .93 in both samples). The six items were used to build the latent variable prosocial behavior. To analyze the measurement model, we conducted confirmatory factor analyses in both samples. Initial analyses suggested the addition of a correlation between the items “is willing to compromise in dispute situations” and “is very tolerant” because the corresponding modification indices were very high (>100) in both samples. This correlation is plausible, as both behaviors indicate that one does not insist on one’s view. Including this correlation, the goodness-of-fit indices indicated that both models fit the data well, χ²_A/B (8/8) = 127.2/91.5, CFI_A/B = .96/.99, TLI_A/B = .92/.97, RMSEA_A/B = .12/.09, and SRMR_A/B = .03/.03. The standardized factor loadings were at least .71_A and .70_B.

General cognitive abilities

The variable general cognitive abilities was assessed with the 46-item short form of the CFT 20 (Grundintelligenztest Skala 2; Weiß, 1998). The CFT 20 uses nonverbal material to measure general fluid ability as conceptualized by Cattell (Weiß, 1998). It measures reasoning ability, which has been shown to have the highest loadings on general intelligence (e.g., Arendasy et al., 2008). The reliabilities of the four raw subtest scores were acceptable (Cronbach’s α = .68 in both samples). We used the subtest scores as indicators to build the latent variable general cognitive abilities.

Self-Concepts

Domain-specific ability self-concepts in mathematics and German were measured with four items, originally developed by Jopt (1978). An example item is: “Nobody’s perfect. I am just not talented in mathematics [German]”. Students rated the items on 4-point scales ranging from1 (agree) to 4 (disagree). Reliabilities for both subjects were good in both samples (Cronbach’s αs ranging from .83 to .86). The four items were used to build the domain-specific ability self-concept latent variables.

Statistical Analysis

We applied structural equation modeling in Mplus Version 8 (Muthén & Muthén, 2017) using the maximum likelihood procedure with non-normality-robust standard errors (MLR). To take the nested data (students within a class) into account, the data were clustered by class, and the “type is complex” option was used. Correlations between the predictors were estimated.

On average, 17.2% of the values of the model variables in Sample A and 13.0% in Sample B were missing (see Table 1 for the proportion of missing data per variable and the section Sample and Procedure for more details). To deal with missing data, we used the full information maximum likelihood (FIML) procedure, a state of the art approach as implemented in Mplus.

We applied three models in a stepwise fashion for each of the four outcomes (i.e., grades and test scores in mathematics and German). Figure 1 shows a visualization of the investigated models. All models included the control variables (i.e., general cognitive abilities, and, corresponding to each outcome, domain-specific self-concept and prior achievement) as predictors. Additionally, Model 1 included externalizing behavior as a predictor of achievement but not prosocial behavior. Model 2 included prosocial behavior as a predictor of achievement but not externalizing behavior. Finally, Model 3 included externalizing and prosocial behavior as predictors of achievement. By analyzing the effects of both behaviors on achievement separately first, we could investigate whether externalizing and prosocial behavior were predictive over and above the control variables. Next, we investigated the unique effects of externalizing and prosocial behavior after considering the extent to which they potentially overlapped.

In a second set of analyses, we investigated whether the relations between social behavior and academic achievement differed for boys and girls in the previously investigated models. For this purpose, we tested gender as a moderating variable by using chi-square difference tests to compare a two-group model where all paths were constrained to be equal between the groups with a model in which the path from externalizing or prosocial behavior to achievement was estimated freely for the groups.

Results

Descriptive Data

The means and standard deviations of all variables are presented in Table 1. In both samples, externalizing behavior was low on average but varied a great deal between students (Range_A = 0 to 1.9; Range_B = 0 to 1.8). The means for prosocial behavior were above the theoretical mean of the scale and covered the whole range of the scale (both ranged from 0 to 5). In both samples and at both times, grades in mathematics and German were satisfactory on average.

Concerning the latent correlations, most variables were significantly correlated (see Table 2). Externalizing behavior was negatively associated with all the other variables, except with mathematics self-concept and German self-concept in Sample B, and its correlations with prosocial behavior were strong in both samples (r_A = −.60, p < .001; r_B = −.47, p < .001). Correlations between the other variables were positive and significant, except for those including mathematics self-concept and German self-concept, which were not always significant.

Table 2.

Latent Bivariate Correlations for Both Samples.

Variable	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Externalizing behavior		−.47***	−.17***	−.16**	−.32***	−.26***	−.31***	−.35***	−.17***	−.20***	−.24***	−.07	−.07
2. Prosocial behavior	−.60***		.23***	.28***	.26***	.27***	.27***	.38***	.20***	.26***	.18***	.09*	.19***
3. Grade math.	−.27***	.28***		.32***	.32***	.23***	.34***	.16***	.34***	.20***	.33***	.33***	−.02
4. Grade German	−.35***	.45***	.52***		.25***	.27***	.25***	.33***	.22***	.33***	.21***	.09*	.24***
5. Test score math.	−.32***	.25***	.45***	.40***		.64***	.66***	.60***	.76***	.62***	.66***	.31***	.19***
6. Test score German	−.20***	.18***	.30***	.36***	.59***		.46***	.53***	.60***	.72***	.52***	.15***	.27***
7. Prior grade math.	−.30***	.31***	.52***	.43***	.47***	.27***		.65***	.57***	.44***	.54***	.31***	.06
8. Prior grade German	−.28***	.40***	.37***	.57***	.31***	.36***	.53***		.49***	.55***	.47***	.08⁺	.30***
9. Prior test score math.	−.22***	.28***	.35***	.33***	.68***	.48***	.39***	.27***		.61***	.70***	.35***	.11**
10. Prior test score German	−.21***	.20***	.32***	.37***	.59***	.71***	.29***	.38***	.52***		.56***	.18***	.28***
11. General cog. abilities	−.18***	.18***	.39***	.31***	.63***	.51***	.43***	.33***	.61***	.58***		.29***	.11*
12. Self-concept math.	−.11**	.12**	.36***	.16***	.30***	.17***	.47***	.15***	.29***	.14***	.28***		.15**
13. Self-concept German	−.10**	.16***	.11*	.28***	.07⁺	.22***	.09*	.37***	.07⁺	.19***	.07	.09⁺

Note. Latent correlations between the variables. The correlations for Sample A are presented below the diagonal. The correlations for Sample B are presented above the diagonal. Math. = mathematics; cog. = cognitive.

⁺p < .10. *p < .05. **p < .01. ***p < .001.

Results of the Main Models

The goodness-of-fit indices indicated that all the structural equation models fit the data well, χ²s (96/95/195) ≤ 293.4/323.6/796.3, CFIs ≥.95, TLIs ≥.94, RMSEAs ≤.05, and SRMRs ≤.05. The exact values for each model are presented in Table 3. The R² values ranged from .14 to .40 for grades and from .53 and .64 for test scores.

Table 3.

Model Fit Indices for the Structural Equation Models for Both Samples.

Fit index	Models for predicting grades in mathematics			Models for predicting grades in German
Fit index	Model 1	Model 2	Model 3	Model 1	Model 2	Model 3
χ² (df)	285.1 (96)	307.3 (95)	785.4 (195)	247.1 (96)	277.3 (95)	752.0 (195)
χ² (df)	275.1 (96)	235.6 (95)	651.1 (195)	275.5 (96)	205.8 (95)	629.3 (195)
CFI	.98	.97	.95	.98	.97	.95
CFI	.98	.98	.97	.98	.99	.97
TLI	.97	.96	.94	.98	.97	.95
TLI	.98	.98	.96	.98	.98	96
RMSEA	.04	.04	.05	.04	.04	.05
RMSEA	.04	.03	.04	.04	.03	.04
SRMR	.03	.03	.05	.02	.02	.05
SRMR	.02	.02	.04	.03	.02	.04
	Models for predicting test scores in mathematics			Models for predicting test scores in German
Fit index	Model 1	Model 2	Model 3	Model 1	Model 2	Model 3
χ² (df)	273.9 (96)	323.6 (95)	796.3 (195)	258.4 (96)	290.5 (95)	769.3 (195)
χ² (df)	275.8 (96)	239.0 (95)	650.4 (195)	293.4 (96)	231.1 (95)	658.9 (195)
CFI	.98	.97	.95	.98	.97	.95
CFI	.98	.99	.97	.98	.98	.97
TLI	.98	.96	.95	.98	.96	.95
TLI	.98	.98	.97	.98	.98	.96
RMSEA	.04	.05	.05	.04	.04	.05
RMSEA	.04	.03	.04	.04	.03	.04
SRMR	.03	.03	.05	.02	.03	.05
SRMR	.02	.02	.04	.03	.03	.04

Note. Model 1 included externalizing behavior as a predictor. Model 2 included prosocial behavior as a predictor. Model 3 included both externalizing and prosocial behavior as predictors. The results for Sample A are presented at the top of each cell. The results for Sample B are presented at the bottom of each cell.

Concerning the models used to predict school grades in mathematics and German, social behavior showed effects on grades over and above the positive effects of the control variables general cognitive abilities, domain-specific self-concepts, and prior grades. See Table 4 for detailed results.

Table 4.

Results From the Structural Equation Models for Both Samples.

Variable	Models for predicting grades in math.						Models for predicting grades in German
	Model 1		Model 2		Model 3		Model 1		Model 2		Model 3
	b*	SE	b*	SE	b*	SE	b*	SE	b*	SE	b*	SE
Externalizing behavior	−.11*	.05	—	—	−.06	.06	−.18**	.05	—	—	−.06	.06
Externalizing behavior	−.05	.04	—	—	.01	.04	−.05	.05	—	—	.02	.05
Prosocial behavior	—	—	.13**	.05	.10⁺	.06	—	—	.26***	.04	.23***	.05
Prosocial behavior	—	—	.15**	.04	.15**	.05	—	—	.17**	.05	.17***	.05
General cog. abilities	.19**	.06	.20**	.06	.19**	.06	.10*	.05	.11*	.05	.11*	.05
General cog. abilities	.18***	.05	.18***	.05	.18***	.05	.07⁺	.04	.07⁺	.04	.08⁺	.04
Self-concept	.13**	.04	.12**	.04	.13**	.04	.08⁺	.04	.07⁺	.04	.08⁺	.04
Self-concept	.23***	.04	.23***	.04	.23***	.04	.16***	.04	.14**	.04	.14**	.04
Prior achievement	.35***	.05	.34***	.05	.33***	.05	.46***	.05	.40***	.04	.39***	.04
Prior achievement	.16***	.04	.14**	.05	.14**	.05	.23***	.06	.20**	.06	.19**	.06
R ²	.33***		.33***		.34***		.37***		.39***		.40***
R ²	.20***		.22***		.22***		.14***		.16***		.16***
	Models for predicting test scores in math.						Models for predicting test scores in German
Variable	Model 1		Model 2		Model 3		Model 1		Model 2		Model 3
Variable	b*	SE	b*	SE	b*	SE	b*	SE	b*	SE	b*	SE
Externalizing behavior	−.16***	.04	—	—	−.14**	.04	−.02	.03	—	—	−.02	.04
Externalizing behavior	−.16***	.03	—	—	−.13**	.04	−.08**	.02	—	—	−.06*	.03
Prosocial behavior	—	—	.09*	.04	.03	.04	—	—	.01	.03	.00	.04
Prosocial behavior	—	—	.12***	.03	.07⁺	.04	—	—	.07**	.03	.06*	.03
General cog. abilities	.31***	.04	.33***	.04	.31***	.04	.14**	.05	.14**	.05	.14**	.05
General cog. abilities	.22***	.05	24***	.05	.22***	.05	.16***	.04	.17***	.04	.16***	.04
Self-concept	.08**	.03	.08*	.03	.08*	.03	.09**	.03	.09**	.03	.09**	.03
Self-concept	.02	.04	.03	.04	.02	.04	.08**	.03	.07**	.03	.07**	.03
Prior achievement	.44***	.04	.44***	.04	.44***	.04	.61***	.05	.61***	.05	.61***	.05
Prior achievement	.57***	.05	.56***	.05	.56***	.05	.60***	.03	.59***	.03	.59***	.03
R ²	.57***		.56***		.57***		.53***		.53***		.53***
R ²	.64***		.62***		.64***		.56***		.56***		.56***

Note. Standardized path coefficients (b*) and standard errors (SE) from the models. Model 1 included externalizing behavior as a predictor. Model 2 included prosocial behavior as a predictor. Model 3 included both externalizing and prosocial behavior as predictors. The results for Sample A (Sample B) are presented at the top (bottom) of each cell. Hyphens indicate that this predictor was not part of the specific model. Math. = mathematics; cog. = cognitive.

⁺p < .10. *p < .05. **p < .01. ***p < .001.

In the first set of models, which included externalizing behavior as a predictor, externalizing behavior had negative effects over and above the effects of the control variables on students’ grades in mathematics and German only in Sample A (b* = −.11, p = .02; b* = −.18, p = .001, respectively). In the second set of models, which included prosocial behavior as a predictor, prosocial behavior had positive effects over and above the effects of the control variables on grades in both subjects in both samples (for grade in mathematics: b*_A = .13, p_A = .01; b*_B = .15, p_B = .001; for grade in German b*_A = .26, p_A < .001; b*_B = .17, p_B = .001). In the third set of models, which included both externalizing and prosocial behavior, externalizing behavior was no longer a significant predictor over and above the effects of the other variables for grades in any subject in any sample. Prosocial behavior was still a positive predictor of the grades in both samples, albeit the effect on grade in mathematics in Sample A was only marginally significant (for grade in mathematics: b*_A = .10, p_A = .09; b*_B = .15, p_B = .001; for grade in German b*_A = .23, p_A < .001; b*_B = .17, p_B < .001).

Concerning the models used to predict test scores in mathematics and German, social behavior also showed effects on the test scores over and above the positive effects of the control variables general cognitive abilities, domain-specific self-concepts, and prior test scores on both test scores (see also Table 4). This is particularly noteworthy as the prior test scores were very strong predictors of subsequent achievement (b*s ranging from .44 to .61, all ps < .001).

In the first set of models, which included externalizing behavior as a predictor, externalizing behavior had negative effects over and above the effects of the control variables on the test scores in mathematics in both samples (b*_A = −.16, p_A < .001; b*_B = −.16, p_B < .001) and on the test score in German in Sample B (b*_B = −.08, p_B = .001). Similarly, in the second set of models, which included prosocial behavior as a predictor, prosocial behavior had positive effects over and above the effects of the control variables on the test scores in mathematics in both samples (b*_A = .09, p_A = .01; b*_B = .12, p_B < .001) and on the test score in German in Sample B (b*_B = .07, p_B = .004). In the third set of models, which included both externalizing and prosocial behavior, externalizing behavior had negative effects over and above the effects of the other variables on the test scores in mathematics in both samples (b*_A = −.14, p_A = .001; b*_B = −.13, p_B = .001). Prosocial behavior only had a marginally significant effect on the test score in mathematics in Sample B (b*_B = .07, p_B = .08). Concerning the test scores in German, externalizing behavior had a negative effect (b*_B = −.06, p_B = .03) and prosocial behavior had a positive effect (b*_B = .06, p_B = .04) only in Sample B.

Gender-Specific Differences

Next, we analyzed whether the relations between social behavior and academic achievement differed between boys and girls. Therefore, we used chi-square difference tests to compare the model with all paths constrained to be equal for both groups with a model in which the paths from externalizing or prosocial behavior to academic achievement were freely estimated for the groups. In the models including externalizing and prosocial behavior (i.e., the third set of models), we ran these tests separately for the effects of externalizing and prosocial behavior on academic achievement. This procedure was used for all of the 24 models we had previously investigated and, because four models per sample included both externalizing and prosocial behavior, for 32 paths from social behavior to achievement. Most models with freely estimated paths did not fit significantly better than the fully constrained models. Nevertheless, in some cases, significant differences occurred.

In Sample A, the model with the freely estimated paths from externalizing behavior to test score in mathematics (Model 1) had a better fit than the fully constrained model (p < .001). The effect was significant for boys (b* = −.15, p = .002) as well as for girls (b* = −.24, p < .001). Further, the model with the freely estimated paths from prosocial behavior to grade in German (Model 2) had a better fit than the fully constrained model (p = .02). Again, the effect was significant for boys (b* = .29, p < .001) and girls (b* = .17, p = .001). Finally, gender differences occurred in the model that included both behaviors to predict test score in mathematics (Model 3). The model with the freely estimated paths from externalizing behavior to the test score in mathematics had a better fit than the fully constrained model (p < .001). The effect was significant for boys (b* = −.11, p = .03) and girls (b* = −.23, p < .001). The model with the freely estimated paths from prosocial behavior to test score in mathematics also had a better fit than the fully constrained model (p = .03). The effect was significant for girls (b* = .10, p = .03) but not for boys (b* = .00, p = .94).

The remaining model comparisons concerning the 12 paths from social behavior to academic achievement in Sample A exhibited no gender differences. Further, in Sample B, all model comparisons concerning the 16 paths from social behavior to academic achievement exhibited no gender differences. See the Supplemental Material for more information on the coefficients for the paths from social behavior to academic achievement in the different models.

Discussion

The main goal of this study was to investigate whether externalizing and prosocial behavior contribute to academic achievement over and above established determinants of academic achievement. Indeed, when both behaviors were analyzed in separate models, we found unique negative effects of externalizing behavior and unique positive effects of prosocial behavior on grades and standardized test scores over and above the other variables. However, when the two behaviors were included in the same model, the result pattern indicated a special relevance of prosocial behavior for grades in both domains and a special relevance of externalizing behavior for test scores in mathematics. Concerning test scores in German, both behaviors had small effects, at least in Sample A. In sum, this study supports prior research that has suggested that externalizing behavior negatively contributes to academic achievement (e.g., van der Ende et al., 2016; Wentzel, 1993; Zhang et al., 2019), further corroborating the adjustment erosion hypothesis (Moilanen et al., 2010). Further, this study is in accordance with research that has suggested that prosocial behavior positively contributes to academic achievement (e.g., Caprara et al., 2000; Curlee et al., 2019; Wentzel, 1993), thus supporting our adjustment accumulation hypothesis. The present study corroborates and expands on prior research because not only did we use a strict longitudinal design and two independent samples, but we also used different measures of achievement, rigorously controlled for the effects of established determinants of academic achievement, and investigated the effects of externalizing and prosocial behavior separately as well as simultaneously in the same models. To the best of our knowledge, there are no studies that have done this before.

Our statistical approach included analyses in which only one social behavior was examined at a time as a predictor of the different types of achievement (i.e., Models 1 and 2) as well as both together (i.e., Model 3). This approach uncovered a very interesting result pattern: When externalizing behavior and prosocial behavior were each analyzed in separate models, both predicted grades and test scores over and above the other determinants of achievement. Interestingly, the result pattern changed when both behaviors were analyzed in the same model. Only prosocial behavior had unique effects on school grades over and above externalizing behavior and the other variables. By contrast, externalizing had a unique effect on test scores in mathematics over and above prosocial behavior and the other variables. Might the ways in which externalizing and prosocial behavior contribute to academic achievement depend on the kind of achievement? The current study at least suggests that this might be the case. For example, the reflection of prosocial behavior in school grades could be due to positive relationships with teachers and classmates and a comfortable learning atmosphere for students exhibiting prosocial behavior, which could encourage active participation in class that should be reflected in school grades (e.g., Caprara et al., 2000; Nurmi, 2012; Weyns et al., 2018). Externalizing behavior may be reflected in test scores in mathematics because, for example, students might miss important content while they are misbehaving. Such a phenomenon can be particularly detrimental in mathematics because the material is systematically built on earlier material and gaps are difficult to compensate for (e.g., Zimmermann et al., 2013). However, these patterns of results need further corroboration.

Additionally, because we included both externalizing and prosocial behavior in our study, our results further support the assumption that they are negatively related but represent different constructs (e.g., Kokko et al., 2006; Padilla-Walker et al., 2018; Schröder et al., 2016). Externalizing and prosocial behavior had strong negative latent bivariate relations (i.e., r_A = −.60 and r_B = −.47, ps < .001), but the correlations were not as high as would be expected for a single construct. Further, they had substantial unique effects on achievement over and above the effect of the respective other behavior, which underpins the assumption that prosocial behavior is more than the mere absence of externalizing behavior and vice versa.

Prior studies have indicated that social behavior might be more relevant for school grades than for test scores (Becherer et al., 2021; Wentzel, 1993; Zimmermann et al., 2013). In our study, the results for the verbal domain also pointed in this direction. However, the results for the mathematical domain did not suggest a special relevance of social behavior for grades compared with test scores. Thus, the findings were inconsistent with regard to achievement measures, and the differences we found should be interpreted with great caution.

We also investigated whether the effects of social behavior on achievement differed by gender, but this was generally not the case. The effects of social behavior on achievement were quite similar for boys and girls, a finding that is consistent with the findings from many other studies (e.g., Deighton et al., 2018; Guo et al., 2018; van der Ende et al., 2016).

Limitations, Future Studies, and Practical Implications

In this study, we investigated samples that were characterized by a rather low socioeconomic status and a large proportion of students with a migration background. Although there was no evidence that these factors influenced our results, the findings should be replicated in samples with different compositions to further investigate the generalizability of our results. Such replication studies should also include both externalizing and prosocial behavior and different achievement measures, as this approach was especially revealing regarding the different contributions that depended on the type of social behavior and achievement measure.

Future studies could also vary the measurement perspective when measuring social behavior to address the robustness of the findings. In this study, we measured social behavior from the teachers’ perspectives. This approach has advantages; for example, more behaviors can be included than would be appropriate with sociometric peer measures, teacher ratings are less influenced by social desirability than self-ratings, because entire schools participated in this study, the teacher ratings were less likely to be influenced by systematic nonparticipation than parent ratings probably would have been and teachers spend much time in the classroom, which gives them many opportunities to compare student behavior. A disadvantage of the teacher ratings used in this study is that report card grades and ratings of social behavior were both from the perspective of teachers (although not necessarily the same teacher). Because the teacher who spent the largest number of hours with the class rated the social behavior, and mathematics and German teachers graded the students (as outcome variable even about 2 years later), it is very unlikely that a single teacher did all the ratings. However, it is possible that the relation between social behavior and school grades would have been smaller if the ratings of social behavior had been made from a different perspective. Future studies could benefit from including different perspectives on students’ social behavior (e.g., observer ratings).

Although we controlled for well-established determinants of school achievement in our study, future studies could include even more covariates of school achievement (e.g., conscientiousness; Meyer et al., 2019; Noftle & Robins, 2007). Further, we focused on the unique effects of social behavior on subsequent academic achievement. However, there is evidence that the relation between externalizing behavior and achievement is reciprocal (e.g., Zhang et al., 2019; Zimmermann et al., 2013). Therefore, future studies should also consider reciprocal dynamics of social behavior and academic achievement.

In our study, externalizing behavior was defined and measured as aggressive and rule-breaking behavior. Besides this widely used narrow definition (e.g., Achenbach et al., 2016; Stanger et al., 1997), externalizing behavior is also defined more broadly and additionally includes symptoms of hyperactivity and inattention (e.g., Hinshaw, 1992; McMahon, 1994). Of course, our results are particularly informative concerning the narrowly defined construct, and it remains an open question whether similar result patterns would have occurred if we had used the broader construct.

A very important question remains: Why does social behavior contribute to academic achievement? The present study’s finding that social behavior contributes to academic achievement over and above established determinants of achievement provides a basis for examining the processes that might underlie these relations. The first studies in this regard have been conducted by now. For example, there is evidence that prosocial behavior is beneficial for students’ relationships with peers and teachers and that through these relationships, prosocial behavior indirectly affects achievement (Becherer et al., 2017; Coulombe & Yates, 2018; Oberle et al., 2022). Concerning externalizing behavior, there is evidence that it is detrimental for achievement via less task-focused behavior (Becherer et al., 2021; Metsäpelto et al., 2015) and that students showing disturbing externalizing behavior in class are given worse grades than their classmates (Krämer & Zimmermann, 2020). These are important first steps for a better understanding of the underlying processes, but more research on mediating processes is needed for a better understanding of the relation(s) between social behavior and (different measures of) achievement.

Such a better understanding of the mechanisms that underlie the relation between social behavior and academic achievement could lead to purposeful and specific practical implications. Nevertheless, at this point, this study’s results further support the need to pay attention to students’ social behavior in everyday school life. The aim should be not only to decrease externalizing behavior but also to promote prosocial behavior. In this regard, students can benefit from programs that address social and emotional learning. For instance, Durlak et al.’s (2011) meta-analysis of more than 200 studies showed that such programs successfully decrease problem behavior and increase positive social behavior and academic achievement.

Conclusion

This study highlights how externalizing and prosocial behavior contribute to academic achievement over and above well-established predictors of achievement. The results indicate that prosocial behavior is particularly relevant for achievement as measured with school grades, whereas externalizing behavior seems more relevant for achievement as measured with standardized test scores in mathematics. The finding that both types of behavior had unique effects when included simultaneously in the analyses further supports that externalizing and prosocial behavior are separate constructs. The findings are very important for predicting secondary school students’ academic outcomes and contribute to a better understanding of the relation that both externalizing and prosocial behavior have with school achievement.

Supplemental Material

Supplemental Material - Does Rude or Kind Behavior Predict Later Academic Achievement? Evidence From Two Samples of Adolescents

Supplemental Material for Does Rude or Kind Behavior Predict Later Academic Achievement? Evidence From Two Samples of Adolescents by Julia Becherer, Olaf Köller, and Friederike Zimmermann in The Journal of Early Adolescence

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 research reported in this article is part of the project “Development and implementation of a school-to-work transition concept for schools serving disadvantaged communities” (“Entwicklung und Implementierung eines neuen Konzeptes zur Eingliederung Jugendlicher in die Berufs-und Arbeitswelt in Schulen mit erhöhtem Förderbedarf” [EIKA]) directed by Olaf Köller (Leibniz Institute for Science and Mathematics Education). The project was funded by the Senator for Education and Science of the Free Hanseatic City of Bremen and was co-financed by the European Regional Development Fund (ERDF).

ORCID iD

Julia Jensen

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Julia Jensen, PhD, is a research associate at the Institute for Psychology of Learning and Instruction (IPL) at Kiel University. Her research interests include whether and why social behavior can contribute to academic achievement and, more generally, heterogeneity and inclusion in the school context.

Olaf Köller, PhD, is the Managing Director of Research at the Leibniz Institute for Science and Mathematics Education (IPN) and Professor of Educational Research at the University of Kiel. His research interests include the role of school environments for individual development, methodological problems in large-scale assessments, and the implementation and evaluation of innovations in schools.

Friederike Zimmermann, PhD, is a professor at the Institute for Psychology of Learning and Instruction (IPL) at Kiel University. Her research interests include, for example, inclusion and heterogeneity in the classroom (particularly along the dimension of social behavior), motivation and self-concept, and teachers’ stereotypes, attitudes, and judgments.

References

Achenbach

T. M.

(1991). Manual for the Teacher’s report form and 1991 profile. University of Vermont.

Achenbach

T. M.

Ivanova

M. Y.

Rescorla

L. A.

Turner

L. V.

Althoff

R. R.

(2016). Internalizing/externalizing problems: Review and recommendations for clinical and research applications. Journal of the American Academy of Child and Adolescent Psychiatry, 55(8), 647–656. https://doi.org/10.1016/j.jaac.2016.05.012

Adams

J. W.

Snowling

M. J.

Hennessy

S. M.

Kind

(1999). Problems of behaviour, reading and arithmetic: Assessments of comorbidity using the Strengths and Difficulties Questionnaire. The British Journal of Educational Psychology, 69(4), 571–585. https://doi.org/10.1348/000709999157905

American Psychological Association (2017). Ethical principles of psychologists and code of conduct. https://www.apa.org/ethics/code/ethics-code-2017.pdf

Arbeitsgruppe Deutsche Child Behavior Checklist (1993). Lehrerfragebogen über das Verhalten von Kindern und Jugendlichen; deutsche Bearbeitung der Teacher's Report Form der Child Behavior Checklist (TRF) [Teacher questionnaire about student behavior: German version of the Teacher’s Report Form of the Child Behavior Checklist (TRF)]. Arbeitsgruppe Kinder-, Jugend- und Familiendiagnostik (KJFD).

Arendasy

M. E.

Hergovich

Sommer

(2008). Investigating the ‘g’-saturation of various stratum-two factors using automatic item generation. Intelligence, 36(6), 574–583. https://doi.org/10.1016/j.intell.2007.11.005

Batson

C. D.

Powell

A. A.

(2003). Altruism and prosocial behavior. In Weiner

I. B.

(Ed), Handbook of psychology (5, pp. 463–484). Wiley.

Becherer

Köller

Zimmermann

(2017). Sozialverhalten und Schulleistungen: Spielt die Beliebtheit in der Klasse eine Rolle? [social behavior and school achievement: Does peer acceptance play a role?]. Zeitschrift Für Erziehungswissenschaft, 20(3), 405–424. https://doi.org/10.1007/s11618-017-0771-1

Becherer

Köller

Zimmermann

(2021). Externalizing behaviour, task-focused behaviour, and academic achievement: An indirect relation? The British Journal of Educational Psychology, 91(1), 27–45. https://doi.org/10.1111/bjep.12347

10.

Burt

S. A.

(2012). How do we optimally conceptualize the heterogeneity within antisocial behavior? An argument for aggressive versus non-aggressive behavioral dimensions. Clinical Psychology Review, 32(4), 263–279. https://doi.org/10.1016/j.cpr.2012.02.006

11.

Caprara

G. V.

Barbaranelli

Pastorelli

Bandura

Zimbardo

P. G.

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

12.

Chen

Lee

Stevenson

H. W.

(1996). Long-term prediction of academic achievement of American, Chinese, and Japanese adolescents. Journal of Educational Psychology, 88(4), 750–759. https://doi.org/10.1037//0022-0663.88.4.750

13.

Coulombe

B. R.

Yates

T. M.

(2018). Prosocial pathways to positive adaptation: The mediating role of teacher-child closeness. Journal of Applied Developmental Psychology, 58, 9–17. https://doi.org/10.1016/j.appdev.2018.08.003

14.

Curlee

A. S.

Aiken

L. S.

Luthar

S. S.

(2019). Middle school peer reputation in high-achieving schools: Ramifications for maladjustment versus competence by age 18. Development and Psychopathology, 31(2), 683–697. https://doi.org/10.1017/S0954579418000275

15.

Deighton

Humphrey

Belsky

Boehnke

Vostanis

Patalay

(2018). Longitudinal pathways between mental health difficulties and academic performance during middle childhood and early adolescence. The British Journal of Developmental Psychology, 36(1), 110–126. https://doi.org/10.1111/bjdp.12218

16.

DeVries

J. M.

Rathmann

Gebhardt

(2018). How does social behavior relate to both grades and achievement scores? Frontiers in Psychology, 9, 857–858. https://doi.org/10.3389/fpsyg.2018.00857

17.

Doctoroff

G. L.

Greer

J. A.

Arnold

D. H.

(2006). The relationship between social behavior and emergent literacy among preschool boys and girls. Journal of Applied Developmental Psychology, 27(1), 1–13. https://doi.org/10.1016/j.appdev.2005.12.003

18.

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(1), 405–432. https://doi.org/10.1111/j.1467-8624.2010.01564.x

19.

Eagly

A. H.

Wood

Diekman

A. B.

(2000). Social role theory of sex differences and similarities: A current appraisal. In Eckes

Trautner

H. M.

(Eds), The developmental social psychology of gender (pp. 123–174). Lawrence Erlbaum Associates Publishers.

20.

EIKA . (2009). Projekt Entwicklung und Implementierung eines neuen Konzeptes zur Eingliederung Jugendlicher in die Berufs- und Arbeitswelt in Schulen mit erhöhtem Förderbedarf (EIKA) [development and implementation of a school-to-work transition concept for schools serving disadvantaged communities (EIKA)]. Humboldt-Universität zu Berlin. Max-Planck-Institut für Bildungsforschung, Friedrich-Alexander-Universität Erlangen Nürnberg, Leibniz-Institut für die. Pädagogik der Naturwissenschaften und Mathematik an der Universität Kiel.

21.

Eisenberg

Miller

P. A.

(1987). The relation of empathy to prosocial and related behaviors. Psychological Bulletin, 101(1), 91–119. https://doi.org/10.1037/0033-2909.101.1.91

22.

Finn

J. D.

Fish

R. M.

Scott

L. A.

(2008). Educational sequelae of high school misbehavior. The Journal of Educational Research, 101(5), 259–274. https://doi.org/10.3200/JOER.101.5.259-274

23.

Ganzeboom

H. B. G.

Treiman

D. J.

(1996). Internationally comparable measures of occupational status for the 1988 international standard classification of occupations. Social Science Research, 25(3), 201–239. https://doi.org/10.1006/ssre.1996.0010

24.

Gerbino

Zuffianò

Eisenberg

Castellani

Luengo Kanacri

B. P.

Pastorelli

Caprara

G. V.

(2018). Adolescents’ prosocial behavior predicts good grades beyond intelligence and personality traits. Journal of Personality, 86(2), 247–260. https://doi.org/10.1111/jopy.12309

25.

Goodman

(1997). The strengths and difficulties questionnaire: A research note. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 38(5), 581–586. https://doi.org/10.1111/j.1469-7610.1997.tb01545.x

26.

Guo

Zhou

Feng

(2018). Pro-social behavior is predictive of academic success via peer acceptance: A study of Chinese primary school children. Learning and Individual Differences, 65, 187–194. https://doi.org/10.1016/j.lindif.2018.05.010

27.

Halonen

Aunola

Ahonen

Nurmi

J.-E.

(2006). The role of learning to read in the development of problem behaviour: A cross-lagged longitudinal study. The British Journal of Educational Psychology, 76(Pt 3), 517–534. https://doi.org/10.1348/000709905X51590

28.

Hattie

(2008). Visible learning: A synthesis of over 800 meta-analyses relating to achievement. Routledge.

29.

Hinshaw

S. P.

(1992). Externalizing behavior problems and academic underachievement in childhood and adolescence: Causal relationships and underlying mechanisms. Psychological Bulletin, 111(1), 127–155. https://doi.org/10.1037/0033-2909.111.1.127

30.

Huang

(2011). Self-concept and academic achievement: A meta-analysis of longitudinal relations. Journal of School Psychology, 49(5), 505–528. https://doi.org/10.1016/j.jsp.2011.07.001

31.

Jopt

U. J.

(1978). Selbstkonzept und Ursachenerklärung in der Schule [Self-concept and attribution in school. Kamp.

32.

Klieme

Artelt

Hartig

Jude

Köller

Prenzel

Schneider

Stanat

(Eds), (2010). PISA 2009: Bilanz nach einem Jahrzehnt [PISA 2009: Review after one decade. Waxmann.

33.

Kokko

Tremblay

R. E.

Lacourse

Nagin

D. S.

Vitaro

(2006). Trajectories of prosocial behavior and physical aggression in middle childhood: Links to adolescent school dropout and physical violence. Journal of Research on Adolescence, 16(3), 403–428. https://doi.org/10.1111/j.1532-7795.2006.00500.x

34.

Krämer

Zimmermann

(2020). Zum Einfluss von störendem Schülerverhalten im Unterricht auf Leistungsbeurteilungen: Explizite Einschätzungen und experimentelle Befunde. Zeitschrift Für Pädagogische Psychologie, 34(2), 99–115. https://doi.org/10.1024/1010-0652/a000250

35.

Kulkarni

Sullivan

A. L.

Kim

(2021). Externalizing behavior problems and low academic achievement: Does a causal relation exist? Educational Psychology Review, 33(3), 915–936. https://doi.org/10.1007/s10648-020-09582-6

36.

Lehmann

R. H.

Gänsfuß

Peek

(1999). And school learning among pupils in the city of Hamburg - grade 7]. Aspekte der Lernausgangslagen und der Lernentwicklung von Schülerinnen und Schülern an Hamburger Schulen - Klassenstufe 7. Hamburger Schulbehörde. [Components of initial achievement.

37.

Lehmann

R. H.

Peek

Gänsfuß

Husfeldt

(2002). Hamburger Schulbehörde. LAU 9: Aspekte der Lernausgangslage und der Lernentwicklung -Klassenstufe 9- [LAU 9: Components of initial achievement and school learning - Grade 9 -].

38.

Longobardi

Prino

L. E.

Marengo

Settanni

(2016). Student-teacher relationships as a protective factor for school adjustment during the transition from middle to high school. Frontiers in Psychology, 7, 1988–1989. https://doi.org/10.3389/fpsyg.2016.01988

39.

Loveland

J. M.

Lounsbury

J. W.

Welsh

Buboltz

W. C.

(2007). The validity of physical aggression in predicting adolescent academic performance. The British Journal of Educational Psychology, 77(Pt 1), 167–176. https://doi.org/10.1348/000709905X79563

40.

McMahon

R. J.

(1994). Diagnosis, assessment, and treatment of externalizing problems in children: The role of longitudinal data. Journal of Consulting and Clinical Psychology, 62(5), 901–917. https://doi.org/10.1037/0022-006X.62.5.901

41.

Metsäpelto

R. L.

Pakarinen

Kiuru

Poikkeus

A. M.

Lerkkanen

M. K.

Nurmi

J. E.

(2015). Developmental dynamics between children’s externalizing problems, task-avoidant behavior, and academic performance in early school years: A 4-year follow-up. Journal of Educational Psychology, 107(1), 246–257. https://doi.org/10.1037/a0037389

42.

Meyer

Fleckenstein

Retelsdorf

Köller

(2019). The relationship of personality traits and different measures of domain-specific achievement in upper secondary education. Learning and Individual Differences, 69, 45–59. https://doi.org/10.1016/j.lindif.2018.11.005

43.

Moilanen

K. L.

Shaw

D. S.

Maxwell

K. L.

(2010). Developmental cascades: Externalizing, internalizing, and academic competence from middle childhood to early adolescence. Development and Psychopathology, 22(3), 635–653. https://doi.org/10.1017/S0954579410000337

44.

Montroy

J. J.

Bowles

R. P.

Skibbe

L. E.

Foster

T. D.

(2014). Social skills and problem behaviors as mediators of the relationship between behavioral self-regulation and academic achievement. Early Childhood Research Quarterly, 29(3), 298–309. https://doi.org/10.1016/j.ecresq.2014.03.002

45.

Muris

Meesters

van den Berg

(2003). The strengths and difficulties questionnaire (SDQ): Further evidence for its reliability and validity in a community sample of dutch children and adolescents. European Child and Adolescent Psychiatry, 12(1), 1–8. https://doi.org/10.1007/s00787-003-0298-2

46.

Muthén

L. K.

Muthén

B. O.

(2017). Mplus (version 8. Muthén and Muthén.

47.

Noftle

E. E.

Robins

R. W.

(2007). Personality predictors of academic outcomes: Big five correlates of GPA and SAT scores. Journal of Personality and Social Psychology, 93(1), 116–130. https://doi.org/10.1037/0022-3514.93.1.116

48.

Nurmi

J. E.

(2012). Students’ characteristics and teacher–child relationships in instruction: A meta-analysis. Educational Research Review, 7(3), 177–197. https://doi.org/10.1016/j.edurev.2012.03.001

49.

Oberle

X. R.

Molyneux

T. M.

(2022). Pathways from prosocial behaviour to emotional health and academic achievement in early adolescence. The Journal of early adolescence. Advance online publication. https://doi.org/10.1177/02724316221113349

50.

OECD (2003). The PISA 2003 assessment framework – mathematics, reading, science and problem solving knowledge and skills.

51.

Okano

Jeon

Crandall

Powell

Riley

(2020). The cascading effects of externalizing behaviors and academic achievement across developmental transitions: Implications for prevention and intervention. Prevention Science: The Official Journal of the Society for Prevention Research, 21(2), 211–221. https://doi.org/10.1007/s11121-019-01055-9

52.

Orpwood

Garden

R. A.

(1998). Assessing mathematics and science literacy In: TIMSS monograph no 4). Pacific Educational Press.

53.

Padilla-Walker

L. M.

Carlo

(2014). The study of prosocial behavior: Past, present, and future. In Padilla-Walker

L. M.

Carlo

(Eds), Prosocial development: A multidimensional approach (pp. 3–16). Oxford University Press.

54.

Padilla-Walker

L. M.

Memmott-Elison

M. K.

Coyne

S. M.

(2018). Associations between prosocial and problem behavior from early to late adolescence. Journal of Youth and Adolescence, 47(5), 961–975. https://doi.org/10.1007/s10964-017-0736-y

55.

Romano

Babchishin

Pagani

L. S.

Kohen

(2010). School readiness and later achievement: Replication and extension using a nationwide Canadian survey. Developmental Psychology, 46(5), 995–1007. https://doi.org/10.1037/a0018880

56.

Roorda

D. L.

Jak

Zee

Oort

F. J.

Koomen

H. M. Y.

(2017). Affective teacher-student relationships and students’ engagement and achievement: A meta-analytic update and test of the mediating role of engagement. School Psychology Review, 46(3), 239–261. https://doi.org/10.17105/SPR-2017-0035.V46-3

57.

Roth

Becker

Romeyke

Schäfer

Domnick

Spinath

F. M.

(2015). Intelligence and school grades: A meta-analysis. Intelligence, 53, 118–137. https://doi.org/10.1016/j.intell.2015.09.002

58.

Schröder

Seehagen

Zmyj

Hebebrand

(2016). Wie du mir, so ich dir? Die Entwicklung von prosozialem Verhalten und der Zusammenhang mit externalisierenden und internalisierenden Auffälligkeiten [“Tit for tat?” The development of prosocial behavior and its relationship to externalizing and internalizing problems]. Zeitschrift für Kinder- und Jugendpsychiatrie und Psychotherapie, 44(1), 31–38. https://doi.org/10.1024/1422-4917/a000396.

59.

Spinath

Harald Freudenthaler

Neubauer

A. C.

(2010). Domain-specific school achievement in boys and girls as predicted by intelligence, personality and motivation. Personality and Individual Differences, 48(4), 481–486. https://doi.org/10.1016/j.paid.2009.11.028

60.

Spinath

F. M.

Harlaar

Plomin

(2006). Predicting school achievement from general cognitive ability, self-perceived ability, and intrinsic value. Intelligence, 34(4), 363–374. https://doi.org/10.1016/j.intell.2005.11.004

61.

Stanger

Achenbach

T. M.

Verhulst

F. C.

(1997). Accelerated longitudinal comparisons of aggressive versus delinquent syndromes. Development and Psychopathology, 9(1), 43–58. https://doi.org/10.1017/s0954579497001053

62.

Stanovich

K. E.

(1986). Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quarterly, 21(4), 360–407. https://doi.org/10.1598/rrq.21.4.1

63.

Thorndike

E. L.

(1920). A constant error in psychological ratings. Journal of Applied Psychology, 4(1), 25–29. https://doi.org/10.1037/h0071663

64.

Tremblay

R. E.

Vitaro

Gagnon

Piché

Royer

(1992). A prosocial scale for the preschool behaviour questionnaire: Concurrent and predictive correlates. International Journal of Behavioral Development, 15(2), 227–245. https://doi.org/10.1177/016502549201500204

65.

van der Ende

Verhulst

F. C.

Tiemeier

(2016). The bidirectional pathways between internalizing and externalizing problems and academic performance from 6 to 18 years. Development and Psychopathology, 28(3), 855–867. https://doi.org/10.1017/S0954579416000353

66.

van Lier

P. A. C.

Koot

H. M.

(2010). Developmental cascades of peer relations and symptoms of externalizing and internalizing problems from kindergarten to fourth-grade elementary school. Development and Psychopathology, 22(3), 569–582. https://doi.org/10.1017/S0954579410000283

67.

van Lier

P. A. C.

Vitaro

Barker

E. D.

Koot

H. M.

Tremblay

R. E.

(2009). Developmental links between trajectories of physical violence, vandalism, theft, and alcohol-drug use from childhood to adolescence. Journal of Abnormal Child Psychology, 37(4), 481–492. https://doi.org/10.1007/s10802-008-9289-6

68.

Vannatta

Gartstein

M. A.

Zeller

Noll

R. B.

(2009). Peer acceptance and social behavior during childhood and adolescence: How important are appearance, athleticism, and academic competence? International Journal of Behavioral Development, 33(4), 303–311. https://doi.org/10.1177/0165025408101275

69.

Veenstra

Lindenberg

Oldehinkel

A. J.

De Winter

A. F.

Verhulst

F. C.

Ormel

(2008). Prosocial and antisocial behavior in preadolescence: Teachers’ and parents’ perceptions of the behavior of girls and boys. International Journal of Behavioral Development, 32(3), 243–251. https://doi.org/10.1177/0165025408089274

70.

Vinnick

L. A.

Erickson

M. T.

(1992). Relationships among accumulated lifetime life events, prosocial skills, and behavior problems in elementary school children. Journal of Child and Family Studies, 1(2), 141–154. https://doi.org/10.1007/bf01321282

71.

Weiß

R. H.

(1998). CFT 20: Grundintelligenztest Skala 2 [CFT 20: General intelligence test scale 2 (4th ed.). Hogrefe.

72.

Wentzel

K. R.

(1993). Does being good make the grade? Social behavior and academic competence in middle school. Journal of Educational Psychology, 85(2), 357–364. https://doi.org/10.1037/0022-0663.85.2.357

73.

Wentzel

K. R.

(2005). Peer relationships, motivation, and academic performance at school. In Elliott

A. J.

Dweck

C. S.

(Eds), Handbook of competence and motivation (pp. 279–296). Guilford Publications.

74.

Wentzel

K. R.

Caldwell

(1997). Friendships, peer acceptance, and group membership: Relations to academic achievement in middle school. Child Development, 68(6), 1198–1209. https://doi.org/10.1111/j.1467-8624.1997.tb01994.x

75.

Weyns

Colpin

De Laet

Engels

Verschueren

(2018). Teacher support, peer acceptance, and engagement in the classroom: A three-wave longitudinal study in late childhood. Journal of Youth and Adolescence, 47(6), 1139–1150. https://doi.org/10.1007/s10964-017-0774-5

76.

Wigfield

Eccles

J. S.

(2000). Expectancy-value theory of achievement motivation. Contemporary Educational Psychology, 25(1), 68–81. https://doi.org/10.1006/ceps.1999.1015

77.

M. L.

Adams

R. J.

Wilson

(1998). ConQuest: Generalized item response modeling software [computer software]. Australian Council for Educational Research.

78.

Zhang

Chen

Deater-Deckard

(2019). Developmental changes in longitudinal associations between academic achievement and psychopathological symptoms from late childhood to middle adolescence. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 60(2), 178–188. https://doi.org/10.1111/jcpp.12927

79.

Zimmermann

Schütte

Taskinen

Köller

(2013). Reciprocal effects between adolescent externalizing problems and measures of achievement. Journal of Educational Psychology, 105(3), 747–761. https://doi.org/10.1037/a0032793

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB