The Meaning of School Grades and Self-Concept for General Self-Esteem in Secondary School Students
A Comparison Between the Physical Fitness and Verbal Domain Using the Internal/External Frame of Reference Model
Abstract
Abstract: This cross-sectional study with 825 students from fifth to ninth grade (Mage = 13.01, SD = 1.30) investigated how self-concept domains (physical fitness and verbal) are associated with general self-esteem. Extending the internal/external frame of reference model, school grades in physical education and German (as a subject) were considered as correlates of the verbal and physical fitness self-concept domains. In addition, considering the well-established empirical evidence on gender differences in the different subdomains, measurement and structural invariance were tested across female and male students. Grades, self-concept, and self-esteem were assessed via questionnaires. Structural equation modeling revealed positive relations between grades and the corresponding self-concept subdomains. Physical self-concept had a stronger association with general self-esteem compared with verbal self-concept. Gender stereotypical differences were found only in the mean values of self-concept subdomains and general self-esteem. The results indicate that school grades and the corresponding self-concepts are strongly associated and that physical self-concept might be a salient and meaningful source of general self-esteem in adolescents.
Zusammenfassung: Das Ziel der vorliegenden querschnittlichen Studie war es, vor dem Hintergrund der Annahmen des Internal/External Frame of Reference Modells die Bedeutung von Schulnoten in Sport und Deutsch und Subdomänen des nicht-akademischen und akademischen Selbstkonzepts (physische Fitness Selbstkonzept und verbale Selbstkonzept) für den generellen Selbstwert zu untersuchen. Geschlechterunterschiede wurden analysiert. Es wurden insgesamt 825 Schülerinnen und Schüler der fünften bis neunten Klasse (Mage = 13.01, SD = 1.30) zu ihren Noten, Selbstkonzept und Selbstwert befragt. Die Ergebnisse eines Strukturgleichungsmodells zeigten positive Zusammenhänge zwischen den Noten und den assoziierten Subdomänen des Selbstkonzepts. Das physische Fitness Selbstkonzept korrelierte stärker mit dem Selbstwert im Vergleich zum verbalen Selbstkonzept. Geschlechtsspezifische Unterschiede fanden sich nur in den Mittelwerten der Selbstkonzept-Subdomänen und dem generellen Selbstwert. Die Ergebnisse deuten darauf hin, dass Noten und die assoziierten Selbstkonzepte stark zusammenhängen und, dass das physische Fitness Selbstkonzept eine wichtige und bedeutsame Quelle für den generellen Selbstwert von Jugendlichen sein kann.
Studies across disciplines and theoretical perspectives demonstrate that a positive self-concept in different subdomains in adolescents is associated with desired outcomes such as global self-esteem (e. g., Arens & Hasselhorn, 2014), which again is a predictor of health-related behavior and health (e. g., Trzesniewski et al., 2006). Thus, acquiring a positive self-concept is crucial to adolescent development (e. g., Harter, 2012). Understanding how self-concept develops and how it affects self-esteem is relevant, as this forms the basis for interventions to promote a positive self-concept and healthy development. This is not only significant for research, but is also of interest especially for professionals in the field of health, education, or politics.
Self-concept is, in very broad terms, defined as a person’s perception of him- or herself (Shavelson et al., 1976; Tietjens et al., 2022). According to Shavelson and colleagues (1976), self-concept is multidimensionally and hierarchically structured. Global self-esteem is depicted as the apex of the self-concept model. It is proposed that global self-esteem is influenced by specific self-concept domains, that is, a person’s perception of competencies or abilities in a specific domain, which is formed through experiences and interpretations of one’s environment. Self-concept can be divided into academic and nonacademic domains. These domains can be further subdivided. While academic self-concept entails one’s perceptions in different school subjects, for example, German (German self-concept), nonacademic self-concept comprises, for instance, perceptions in the physical domain (physical self-concept). Although physical education is taught as a school subject, physical self-concept is considered to be a nonacademic self-concept domain. It can be argued that physical activity and sports are to a large extent leisure time activities and that a student’s self-concept is influenced by various sources inside and outside school (e. g., comparison with peers in sports clubs, trainers’ feedback; cf. also Arens & Preckel, 2018; Burrmann, 2015).
Academic and nonacademic self-concept domains, in turn, are influenced by more specific subdomains at the lower level reflecting perceptions of behavior in context-related situations (Scalas et al., 2017). German self-concept, for example, can be split into listening, writing, reading, and speaking. Physical self-concept entails perceptions of physical appearance and physical fitness. The latter can be further differentiated into strength, endurance, speed, flexibility, and coordination. While self-esteem is proposed to be relatively stable, self-concept subdomains depend on specific situations and thus are less stable. Consequently, subdomains, not global self-esteem, are taken into account when investigating influencing factors and developing interventions (Burrmann, 2016; Rubeli, Oswald, Conzelmann, Schmid, et al., 2020).
Influences on Physical Self-Concept
Especially in the physical domain, there is much interest in understanding the effects that physical activity, exercise, and performance have on self-concept. Based on the model by Shavelson and colleagues (1976), Sonstroem and Morgan (1989), for example, developed the exercise and self-esteem model (EXSE-Model), which offers a theoretical framework helping to understand the processes by which physical self-concept and self-esteem are influenced. The model proposes that physical activities and physical exercise influence physical self-efficacy, which is defined as the perceived capability to master specific tasks in the physical domain (Bandura, 1997). The feeling of being self-effective increases physical self-concept (here: perceived physical competence). The positive perceptions of physical competence, in turn, positively influence global self-esteem. The model has been tested and validated in different populations, including adolescents (Rubeli, Oswald, Conzelmann, & Schmidt, 2020; Schmidt et al., 2015), primary school children (Noordstar et al., 2016), and athletes (Sonstroem et al., 1994).
Physical self-concept in adolescents is reported to be influenced not only by the number of physical activities and physical exercises, but also by actual physical performance. For example, Schmidt et al. (2015) showed that physical self-concept is influenced by the results of a motor abilities test. Moreover, results revealed that physical self-concept is a crucial mediator in the association between motor abilities and self-esteem in adolescents. Another factor that may influence physical self-concept, especially in adolescents, is school grades. Grades are easy to compare, and thus they enable students to compare themselves with their peers, and also to compare their competencies in different subjects. The internal/external frame of reference model (I/E model; Marsh, 1986), which is one of the most validated models in self-concept research, offers a theoretically and empirically testable framework to explain the influence of achievement (e. g., expressed by grades) on the development of self-concept subdomains.
The Internal/External Frame of Reference Model
According to the I/E model, the development of self-concept subdomains relies on different achievement comparison processes, that is, social, dimensional, and temporal comparisons. In social comparison processes, students compare their grades in one particular subject with the grades of others (e. g., their classmates; external frame of reference). It can be assumed that students with a lower grade in one particular domain compared to the grade of their classmates will develop a lower self-concept in that domain. At the same time, in dimensional comparison processes, students compare their grades in one particular domain with those in other domains (internal frame of reference). This can lead to a decreased self-concept in the subdomain that is associated with the worse grade (contrast effects). In dimensional comparison processes, students may also compare the effort they put into tasks in different subdomains (e. g., to reach goals or master tasks). Temporal comparisons in which students compare their achievement in a subject with their prior achievement in this subject have been recognized as a third process influencing self-concept formation (Möller & Trautwein, 2020; Wolff et al., 2019).
Originally, the I/E model was restricted to the math and verbal domain. Later, in particular Möller and colleagues extended the I/E model by investigating effects adding further self-concept domains, such as social sciences (i. e., history and politics) or natural sciences (i. e., biology, chemistry, and physics; Jansen et al., 2015; Möller et al., 2016). Studies showed that the dimensional comparison effects depend on the compared self-concept subdomains. It is reported that strong contrast effects only exist in contrasting subdomains that are on opposite ends of the theoretical continuum of (academic) self-concept (i. e., math and verbal), but much weaker negative effects or even assimilation effects occur for subdomains that are perceived as similar (i. e., math and physics; Jansen et al., 2015). Therefore, it was concluded “that the dimensional comparison effects are moderated by the perceived similarity of the compared school subjects” (Möller et al., 2016, p. 4). As a result, the generalized internal/external frame of reference model (GI/E; Möller et al., 2016) was introduced, which should serve as a guide to look for more I/E relations between independent and dependent variables. As stated by Möller and Marsh (2013), the extension to the nonacademic self-concept subdomains (i. e., sports, music, and arts) seems to be promising, as this would lead to a more comprehensive understanding of the development of self-concept.
Including the Physical Domain in theI/E Model
To date, research investigating the I/E model mainly focuses on the academic domain, and research in the nonacademic domain such as the physical domain is largely missing, being focused on the external frame of reference (e. g., Chanal et al., 2005; Trautwein et al., 2008) or the reciprocal effect model (e. g., Garn et al., 2019; Marsh et al., 2007).
Exceptions are three studies, which were conducted by Tietjens et al. (2005), Lohbeck et al. (2021), and Arens and Preckel (2018). While Tietjens et al. (2005) and Lohbeck et al. (2021) investigated effects on sports-specific self-concepts (e. g., swimming), Arens and Preckel (2018) focused on the physical ability self-concept. The consideration of the physical appearance self-concept would not fit the idea of the I/E model, which is to investigate the influence of achievement in a particular domain on self-concepts (What is achievement in the physical appearance domain?).
Tietjens et al. (2005) extended the I/E model to the investigation of achievement and physical self-concept in two individual sports (i. e., swimming and athletics) and two team sports (i. e., soccer and basketball) in university students. They found positive associations between performance and the corresponding self-concept in each sport. Structural equation modeling showed some negative relations between performances and noncorresponding self-concepts. In 2012, Lohbeck and colleagues (2021) expanded this research by considering six different sports (three individual and three team sports). Their results confirmed the findings by Tietjens et al. (2005).
In a cross-sectional study, Arens and Preckel (2018) investigated the relationships between primary school children’s school grades in the subjects math, German, and physical education with their math, German and physical ability self-concepts (third grade; mean age = 8.90 years).
Moreover, associations between the children’s school grades and their intrinsic values related to math, German, and physical abilities were examined (expectancy-value theory; Eccles & Wigfield, 1995). The findings replicated the assumptions of the original I/E model by showing positive relations between the children’s school grades in math and their math self-concept. Furthermore, the children’s school grades in the subject German were associated with their German self-concept. Negative relations between school grades and noncorresponding self-concept subdomains were found, that is, between the school grade in the subject German and math self-concept (β = −.26, p < .05) as well as between the school grade in math and German self-concept (β = −.28, p < .05). Furthermore, anticipated positive correlations between school grades in physical education and physical self-concept and physical intrinsic value, respectively, were shown. Negative relations were found between math grades and physical ability self-concept (β = −.12, p < .05) and between grades in the subject German and physical intrinsic value (β = −.16, p < .05).
To our knowledge, the study by Arens and Preckel (2018) is the first study testing the I/E model across the academic and nonacademic domains. Their results partly confirmed the negative relations that were assumed between school grades and the noncorresponding self-concepts and intrinsic values. It is concluded that primary students are already able to compare their school grades across the academic and nonacademic domains. However, the inconsistent results might be attributable to developmental processes. The cognitive maturity that allows comparisons to be made across the academic and nonacademic domains might develop from this age onward. Further studies including older students are needed to get further insight into how grades are associated with the academic and nonacademic self-concepts in different age groups. The results might help to understand the use of social and dimensional comparison processes across the academic and nonacademic domains in younger versus older students. Moreover, there is still a need for studies examining the relationship between the academic and non-academic domains with self-esteem in adolescence considering the physical self-concept as part of the nonacademic domain (Arens & Hasselhorn, 2014; Lindwall et al., 2011). The results of these studies could provide insight into whether physical self-concept might be such a meaningful source for general self-esteem as is assumed.
Consideration of Gender Differences
The empirical evidence on gender differences in the self-concept and self-esteem of adolescents is unambiguous. Studies have shown gender-stereotypic differences, that is, in the verbal domain female students have a higher self-concept, whereas in the physical fitness domain male students report a higher self-concept (Esnaola et al., 2018, (2020). In addition, gender differences have also been found in self-esteem, with males consistently reporting higher self-esteem than females. As shown in a meta-analysis, gender differences in self-esteem increase during childhood and peak in adolescence (Zuckerman et al., 2016). Female students’ self-esteem has also been shown to fluctuate more during adolescence (Helwig & Ruprecht, 2017). Explanations for the gender differences in self-concept and self-esteem often relate to gender-stereotypic socialization processes (Eccles et al., 1993). While girls have been found to assign higher importance to verbal academic skills, boys have been found to have higher importance rankings for physical fitness (Burrmann, 2015; Jacobs et al., 2002).
Based on these differences, one can assume that gender differences also occur in the relationship between self-concept and self-esteem. Taking into account correlations between self-concept and self-esteem in female versus male students, Coelho et al. (2020) found a stronger association between academic self-concept and self-esteem in fourth-grade girls (r = .70, p < .001) compared to fourth-grade boys (r = .63, p < .001). By contrast, Arens and Hasselhorn (2014) reported slightly stronger correlations between German self-concept and self-esteem, math self-concept and self-esteem, and physical ability self-concept and self-esteem in fifth- and sixth-grade boys compared to girls. However, in invariance analyses, they did not find any gender differences in self-concept–self-esteem relations. Rubeli and colleagues (Rubeli, Oswald, Conzelmann, & Schmidt, 2020) focused on the relationship between perceived sports competence and self-esteem in female and male adolescents aged 8 – 14 years. In their study, the results of multigroup analyses revealed differences in the mean level of the constructs, but no gender differences with regard to the relations between perceived sports competence and self-esteem.
Research focusing on gender differences in the I/E relations, that is, between achievement and self-concept, has yielded inconsistent results. While several studies demonstrated invariance across gender (e. g., Arens et al., 2017; Möller et al., 2011), some found contradictory results. For example, Skaalvik and Rankin (1990) showed stronger positive effects of verbal achievement on verbal self-concept in boys and a stronger negative effect of verbal achievement on math self-concept in girls. The results of a study by Nagy et al. (2006) demonstrated a negative effect of achievement in math on biology self-concept in boys only (for an overview, see Wolff & Möller, 2022). Only Arens and Preckel (2018) analyzed gender differences in the relationship between achievement and self-concepts across the academic and nonacademic domains including the physical domain. They focused on primary school children. Conducting invariance analyses, they found the relations between the children’s school grades (in the subjects German, math, and physical education) and their self-concepts (German self-concept, math self-concept, physical ability self-concept) to be invariant across gender. The investigation of gender differences in I/E relations across the academic and nonacademic domains in older students remains a research gap.
The Present Investigation
Physical self-concept seems to play an important role in general self-esteem across adolescence. However, there is still a lack of evidence on the role of physical self-concept in contrast to other self-concept domains. Also, it is still not clear how school grades in physical education and other subjects (as comparisons) impact physical self-concept. Finally, there is still an open research questions on whether the meaning of school grades on physical self-concept and the impact of physical self-concept on self-esteem are similar in female and male students.
The objective of the present study is threefold. First, the aim of this study is to investigate the associations between physical self-concept (as a subdomain of the nonacademic domain of self-concept) and verbal self-concept (as a subdomain of the academic domain of self-concept) with general self-esteem (Figure 1). We hypothesize that stronger associations exist between physical self-concept and self-esteem compared to verbal self-concept and self-esteem. Second, considering the I/E model, another aim of this study is to analyze the relations between school grades and self-concept in two subdomains of the academic and nonacademic domains (physical fitness and verbal domain) in adolescent boys and girls. With this, we aim to extend current knowledge on the I/E model across the academic and nonacademic domains in adolescent students. We hypothesize that positive relations will be found between school grades and the corresponding self-concept subdomains. In addition, negative correlations between the grades and noncorresponding self-concepts are hypothesized.
Third, taking into account the well-established empirical evidence on gender differences in the different subdomains of the self-concept, the last aim of this study is to test for measurement and structural invariance across gender.
Method
Participants and Procedure
The sample of this cross-sectional study is self-selected and comprises students from fifth to ninth grade (N = 825, Mage = 13.01, SD = 1.30, 58 % female) from four secondary schools in North Rhine-Westphalia, Germany (5.6 % 5th grade, 21.8 % 6th grade, 33.6 % 7th grade, 36.7 % 8th grade, 2.3 % 9th grade). For fifth- and sixth-grade students, written informed consent was obtained from a parent or legal guardian. According to the school law in North Rhine-Westphalia (§120(2)), older students from grade 7 on were able to provide consent for themselves. The survey was conducted in a classroom setting. After a short introduction by a trained research assistant, the students filled out the questionnaire on their own. Questions were related to the student’s grades, self-concept, and self-esteem. As the students were asked to report their latest grades on their reports, the assignment of the grade was before students filled out the survey on self-concept and self-esteem. In the case of ambiguities or difficulties in understanding, the students could ask for individual help. Students took part voluntarily and could stop without reason at any time. The study and consent procedure were approved by the university ethics committee.
Measures
Physical fitness self-concept
Physical fitness self-concept was assessed using the physical fitness scales of the physical self-concept questionnaire by Stiller and colleagues (2004; PSK). Each domain (strength, endurance, speed, coordination, flexibility, and sportiness) of physical fitness self-concept was measured using six items (e. g., I am strong [Ich bin stark], 36 items in total). Responses were reported on a 4-point Likert scale (1 = totally disagree to 4 = totally agree). A mean score (total physical fitness self-concept) was formed considering all domains.
Verbal self-concept
Verbal self-concept was assessed using four skill-specific scales (speaking, listening, writing, and reading) and one global scale related to the student’s native language (e. g., Reading texts is easy for me [Das Lesen von Texten fällt mir leicht]). In this study, the native language of the students was German. Therefore, the verbal self-concept related to the German language was assessed. Each scale consists of three items on a 4-point Likert scale (1 = totally disagree to 4 = totally agree, 15 items in total; Arens & Jansen, 2016). A mean score (total verbal self-concept) was formed from the five domains.
Self-esteem
Self-esteem was measured by eight items of a German version of the Self-Description Questionnaire I (Marsh, 1990b; adapted by Arens et al., 2011; e. g., I do lots of important things [Ich mache viele wichtige Dinge]) on a 5-point Likert scale (1 = totally disagree to 5 = totally agree). A mean score was calculated considering the eight items.
Achievement
In the German secondary school system, teachers give grades for each subject in half-year intervals considering the knowledge gained in the class, competencies learned, and abilities acquired. Grades range from 1 (highest achievement) to 6 (lowest achievement) with possible graduations of plus (+) or minus (–) Students were asked for their grades in the subjects physical education and German from the latest school reports.
Data Analysis
Data analyses were conducted with SPSS (v. 26) and RStudio (v. 1.2.5033). Means and standard deviations of each subdomain were calculated. In addition, reliability scores were computed. Based on the literature (i. e., Arens and Morin, 2016), it can be presumed that the self-concept items load differently weighted on their associated latent factor. Therefore, it is recommended to calculate reliability scores based on a congeneric model (McDonald’s omega) instead of an equivalent tau measurement model (Cronbach’s alpha). Correlations between subdomains were analyzed using Spearman’s rho (nonnormality distribution of ordinal data; Kolmogorov–Smirnov test, p < .05). According to Cohen (1988; Sedlmeier & Renkewitz, 2008), correlation coefficients of r = .1 indicate a weak association between variables. Moderate associations can be assumed when r = .3, and strong associations when r = .5.
Next, structural equation modeling was conducted to investigate the relations between school grades and self-concept in the academic and nonacademic domains, as well as its associations with self-esteem (Figure 1). As the variables are ordinally scaled and not normally distributed, a diagonally weighted least squares estimator (DWLS) was chosen, which is specifically designed for nonnormal ordinal data with a small number of categories (Mîndrila, 2010). Missing values were removed listwise. Commonly used methods to evaluate model fits are used: χ2 statistics, the comparative fit index (CFI), the Tucker–Lewis index (TLI), and the root mean square error of approximation (RMSEA), which is based on the χ2 statistics (Shi et al., 2018). It is stated that CFI and TLI values greater than .95 and RMSEA values lower than .08 indicate a good fit of the data (Hu & Bentler, 1999).
The research question regarding gender differences in the I/E model pertains to the concept of structural invariance, which should be only taken into account if measurement invariance can be identified a priori (cf. Vandenberg, 2002). Therefore, in the present study, we used the equaltestMi package (version 0.6.0) to test measurement and structural invariance stepwise in a way that is well established (i. e., Putnick & Bornstein, 2016; Vandenberg, 2002). We made use of several model constraints resulting in nested models. In the first model, we tested whether factor loadings are invariant across gender (metric model). In the second model, item intercepts were additionally constrained (scalar model). As the last step in testing measurement invariance, residuals were additionally constrained (strict model). Each model was compared to the baseline model using descriptive goodness-of-fit indices. According to Cheung and Rensvold (2002) and Chen (2007), the models compared are assumed to be invariant unless CFI does not differ more than .01. To analyze whether the invariant relations occur across gender, the factor variances and covariances were additionally constrained in Models 4 and 5 (structural invariance). Again, these models were compared to the baseline model using descriptive goodness-of-fit-indices (Δ CFI). Finally, the mean values of the latent constructs were restricted to investigate whether differences occur in the mean levels in female and male students (Model 6).
Results
Results of pre-analyses reveal that most of the students in our sample (physical education: 53 %; German: 40 %) received the grade 2 (“good”). The distributions of the grades in the subjects physical education and German differ significantly (χ2 (156, N = 787) = 388.92, p < .001, V = .20). An overview is presented in Table 1.
Means (M), standard deviations (SD), reliability scores, and correlations between the subdomains of self-concept (i. e., strength, endurance and reading, listening) and self-esteem are presented in Table 2. All scales show at least acceptable reliability. Correlations between subdomains within physical fitness self-concept and verbal self-concept seem to be moderate to strong. The relationship between general self-esteem and physical fitness self-concept is stronger than the relationship between verbal self-concept and general self-esteem. Means of verbal self-concept are higher than those of physical fitness self-concept (except writing).
The results of the structural equation model are presented in Figure 2. A total of 697 students were included in the model. The model shows an excellent model fit (χ2 = 360.69; df = 183, p < .05; CFI = .986; TLI = .984; RMSEA = .036). The correlation between the grades in the subjects physical education and German is significant. The same applies to the correlation between both self-concept domains. As presumed, both the grade in physical education and the grade in the subject German are related to the corresponding self-concept. Furthermore, a weak negative association is found between the grade in physical education and verbal self-concept. No significant relation is found between the grade in the subject German and physical fitness self-concept.
Furthermore, findings indicate that the association between physical fitness self-concept and self-esteem is stronger than the association between verbal self-concept and self-esteem. Results show that 19 % of the variance of physical fitness self-concept and 24 % of verbal self-concept could be explained in this model. Even 45 % could be explained with regard to general self-esteem.
Results indicate strict measurement invariance across gender, as the CFI value does not decrease substantially constraining factor loadings (Model 1), item intercepts (Model 2), and residuals (Model 3). Furthermore, the invariance of factor variances (Model 4) and covariances (Model 5) can be seen as established (see Table 3). The model fit declines more substantially when invariant factor means are additionally assumed. Further testing shows that male students rate themselves higher in physical fitness self-concept (Mfemale = 2.88, Mmale = 3.06) and general self-esteem (Mfemale = 3.88, Mmale = 4.02), whereas female students rate themselves higher in verbal self-concept (Mfemale = 3.24, Mmale = 3.11).
Discussion
In this investigation, we aimed to analyze the associations between physical fitness self-concept (as a subdomain of nonacademic self-concept), verbal self-concept (as a subdomain of academic self-concept), and general self-esteem. Overall, a stronger association exists between physical fitness self-concept and self-esteem compared to the association between verbal self-concept and self-esteem, which is still moderate. This might be an indication that physical fitness self-concept and verbal self-concept are related to self-esteem differently. The stronger correlation between physical self-concept and self-esteem can be also linked to the question of which subdomains are more salient or individually important during adolescence. As physical self-concept research shows, physical self-concept is not only affected by physical education lessons (e. g., in the ASDQII, Marsh, 1990b) but also in a more comprehensive approach by physical activity behavior in general (e. g., Marsh & Redmayne, 1994). It could be argued that sport and physical activity represent the more salient source of self-esteem than school performance in single subjects. Therefore, the related self-concept subdomain is more important in contributing to general self-esteem (cf. research on self-efficacy, e. g., Butz & Usher, 2015). This seems to be reasonable as research shows that the physical self-concept domain is not affected that much by grades compared with other domains that are more directly related to school subjects such as math or language (Marsh et al., 2007). Therefore, the sources for the physical self-concept are more complex because this domain is affected not only by school experiences but also by daily activities such as those in sport clubs, in extra-curricular physical activity, and daily habits. The results reported by Herrmann and colleagues (Herrmann et al., 2014) highlighting that motor test results correlate higher than grades with physical self-concept underline this assumption: The physical self-concept is built by many different activities and habits of children and adolescents. Thus, it can be assumed that it also plays a more important role for general self-esteem as it is more salient for children in their daily life. Also, the physical domain could underlie more social comparisons in the daily life of adolescents than any academic domain (Trautwein et al., 2008). Since these social comparisons might influence not only self-perception itself but also the importance of the domain, this could lead to a higher impact of the physical subdomain on general self-esteem (Van der Aar et al., 2018).
Recent studies by Sohnsmeyer and Heim (2015) and by Rubeli, Oswald, Conzelmann, and Schmidt (2020) investigated whether differences in the relationship between physical self-concept and self-esteem exist according to the individual’s importance of sports competence (individual importance hypothesis). While Rubeli, Oswald, Conzelmann, and Schmidt (2020) showed that the association between physical self-concept and self-esteem was moderated by the individual’s importance of sports competence, interestingly, Sohnsmeyer and Heim (2015) could not confirm these results. Reasons might be the different assessments of physical self-concept and self-esteem or the different analyses strategies. Further research should also focus on the individual importance hypothesis more closely.
Second, we analyzed the relationship between grades and self-concept in the two different domains (i. e., verbal and physical fitness self-concept). We hypothesized positive associations between grades and the corresponding self-concept subdomains and negative associations between grades in physical education and verbal self-concept and between grades in the subject German and physical self-concept. Our analyses showed that school grades are significantly related to the corresponding self-concept subdomains. Although the two analyzed subdomains, verbal and physical fitness, are said not to be related to each other as they belong to the academic and nonacademic domain, respectively (Shavelson et al., 1976), we found small correlations between both grades in the two subjects and the self-concept subdomains. Referring to the I/E relations, the results show only a weak negative relationship between the grade in the subject German and physical fitness self-concept. The association between the grade in physical education and verbal self-concept is not significant. This does not meet the assumption made by Marsh (1986) and Möller et al. (2016), stating that these two different subdomains should lead to contrasting effects. Although a weak negative association between the grade in the subject German and physical fitness self-concept is shown, this is not in line with the results of Arens and Preckel (2018), who did not find a significant relationship between the grade in physical education and verbal self-concept, but did find an association between the grade in math and physical self-concept. Thus, inconsistent results might not be attributable to cognitive maturity as stated before. Instead, since correlations between the two grades were also found, further research should focus on the question of whether these subdomains are as unrelated as proposed in youth. However, it could also be that the correlation is based on a third variable such as participation in class or social behavior, since school grades are not exclusively based on performance measurement.
Considering the well-established empirical evidence on gender differences in the different subdomains of self-concept, measurement invariance and structural invariance across female and male students were analyzed. The results indicate that female and male students do not differ with regard to the relations between constructs. However, they differ in the level of self-concept subdomains, where boys rated themselves higher in physical fitness self-concept and girls in verbal self-concept. These differences could have arisen due to gender differences in the actual physical fitness and school performances in the specific school subjects. The stereotypic differences are in line with previous research (e. g., Marsh et al., 2017).
Limitations
This study is not without limitations. First, data must be interpreted carefully as only cross-sectional results are presented. Second, this study included significantly more girls than boys, which influenced the test for potential group differences, where data were split by gender. Future research should similarly consider girls and boys, as they might have different perceptions of the importance of the subjects physical education (or sport in general) and German (or native language). Third, this study took only one subdimension of the academic and the nonacademic self-concept into account. This leads to the question of whether the same results could have been shown including more than one or other subdimensions of the academic and nonacademic self-concept. The information regarding school grades was generated by self-reports of the children. Although children in this age group are able to report their marks correctly, single reporting biases cannot be excluded totally. Finally, the assessment of the SDQ needs to be discussed. In this study, a German version of the SDQ I was used to assess self-esteem. The version was validated and tested with children from Grades 3 to 6 (Arens, 2011). Marsh and colleagues (Marsh, 1990a) developed a second version of the SDQ, which is validated for adolescents from Grades 7 to 10 (approximately 12 – 17 years). As our sample also included fifth- and sixth-graders, we decided to stay with the SDQ I. The scale showed very good reliability scores for that sample.
Conclusion
The results presented here are a substantial contribution to the current status quo of research regarding the differentiation of self-concept in early adolescence. First, we find significant relations between self-concept subdomains and general self-esteem in this particular age group. This raises the question as to whether bottom–up processes of specific self-concept subdomains can influence general self-esteem levels in adolescence. Moreover, the association seems to be stronger for the physical fitness self-concept domain than for the verbal self-concept subdomain. This hints at the possibility of fostering self-esteem through physical activities and emphasizes the importance of physical activity, physical education, and sport in this age group from a psychological perspective.
Second, grades seem to be strongly related to the level of the corresponding self-concept subdomains. This links to the general discussions about the necessity and sense of grades in specific school settings (Hattie & Timperley, 2007). The possible emphasis on social comparison through school grades could lead to a motivational downward spiral in weaker students. Since negative associations between grades and self-concept subdomains found are weaker than proposed, the importance of factors within the specific domain (e. g., grades, salience by the amount of time invested in, etc.) are hereby emphasized.
Third, differences found for gender are in line with well-established empirical evidence. Summarizing these results, further research should focus more on the importance of and the correlations between different school grades and their corresponding self-concept subdomains. Specifically for interventions concentrating on fostering general self-esteem, the results are helpful as they show that those interventions could (a) focus on different subdomains as they might also influence general self-esteem in a kind of bottom-up processes; (b) differentiate between boys and girls in setting and procedure; and (c) consider physical activity and sport as a potential source, not only for physical self-concept but also for general self-esteem taking into account its salience and importance within the social environment in youth and adolescence.
Literatur
(2017). Social and dimensional comparisons in math and verbal test anxiety: Within-and cross-domain relations with achievement and the mediating role of academic self-concept. Contemporary Educational Psychology, 51, 240 – 252. https://doi.org/10.1016/j.cedpsych.2017.08.005
(2014). Age and gender differences in the relation between self-concept facets and self-esteem. The Journal of Early Adolescence, 34 (6), 760 – 791. https://doi.org/10.1177/0272431613503216
(2016). Self-concepts in reading, writing, listening, and speaking: A multidimensional and hierarchical structure and its generalizability across native and foreign languages. Journal of Educational Psychology, 108 (5), 646 – 664. https://doi.org/10.1037/edu0000081
(2016). Examination of the structure and grade-related differentiation of multidimensional self-concept instruments for children using ESEM. The Journal of Experimental Education, 84 (2), 330 – 355. https://doi.org/10.1080/00220973.2014.999187
(2018). Testing the internal/external frame of reference model with elementary school children: Extension to physical ability and intrinsic value. Contemporary Educational Psychology, 54, 199 – 211. https://doi.org/10.1016/j.cedpsych.2018.06.003
(2011). Erfassung des Selbstkonzepts im mittleren Kindesalter: Validierung einer deutschen Version des SDQ I [Self-Concept Measurement with Preadolescent Children: Validation of a German Version of the SDQ I]. Zeitschrift für Pädagogische Psychologie, 25 (2), 131 – 144. https://doi.org/10.1024/1010-0652/a000030
(2011). Selbstkonzepte von Schülern der Klassenstiufen 3 bis 6: Messung und Validierung der multidimenionalen Struktur [Measurement and Validation of the Multidimensional Self-concept Structure of German Students attending Grades 3 to 6]. Georg-August-Universität Göttingen. https://doi.org/10.53846/goediss-479
(1997). Self-efficacy: The exercise of control. New York: Freeman.
(2015). Effekte des Sporttreibens auf die Entwicklung des Selbstkonzepts Jugendlicher [The effects of sports activities on the self-concept of adolescents]. Zeitschrift für Sportpsychologie, 11 (2), 71 – 82. https://doi.org/10.1026/1612-5010.11.2.71
(2016). Die Stärkung des Selbstkonzepts von Heranwachsenden durch Sport!? [Strengthening the self-concept of adolescents through sports!?]. Bewegung & Sport, 70 (1), 18 – 23.
(2015). Salient sources of early adolescents’ self-efficacy in two domains. Contemporary Educational Psychology, 42, 49 – 61. https://doi.org/10.1016/j.cedpsych.2015.04.001
(2005). Big-fish-little-pond effects on gymnastics self-concept: Social comparison processes in a physical setting. Journal of Sport and Exercise Psychology, 27 (1), 53 – 70. https://doi.org/10.1123/jsep.27.1.53
(2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 14 (3), 464 – 504. https://doi.org/10.1080/10705510701301834
(2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 9 (2), 233 – 255. https://doi.org/10.1207/S15328007SEM0902_5
(2020). A multilevel analysis of the importance of school climate for the trajectories of students’ self-concept and self-esteem throughout the middle school transition. Journal of Youth and Adolescence, 49 (9), 1793 – 1804. https://doi.org/10.1007/s10964-020-01245-7
(1988). Statistical power analysis for the behavioral sciences. Lawrence Earlbaum Associates.
(1995). In the mind of the actor: The structure of adolescents’ achievement task values and expectancy-related beliefs. Personality and Social Psychology Bulletin, 21 (3), 215 – 225. https://doi.org/10.1177/0146167295213003
(1993). Age and gender differences in children’s self‐and task perceptions during elementary school. Child Development, 64 (3), 830 – 847. https://doi.org/10.1111/j.1467-8624.1993.tb02946.x
(2018). Internal structure of academic self-concept through the Self-Description Questionnaire II-Short (SDQII-S). Learning and Individual Differences, 62, 174 – 179. https://doi.org/10.1016/j.lindif.2018.02.006
(2020). The development of multiple self‐concept dimensions during adolescence. Journal of Research on Adolescence, 30(S1), 100 – 114. https://doi.org/10.1111/jora.12451
(2019). Reciprocal effects model of children’s physical activity, physical self-concept, and enjoyment. Psychology of Sport and Exercise, 45, 101568. https://doi.org/10.1016/j.psychsport.2019.101568
(2012). The construction of the self: Developmental and sociocultural foundations (2nd ed.). Guilford Publications.
(2007). The power of feedback. Review of Educational Research, 77 (1), 81 – 112. https://doi.org/10.3102/003465430298487
(2017). Age, gender, and self-esteem: A sociocultural look through a nonparametric lens. Archives of Scientific Psychology, 5 (1), 19 – 31. https://doi.org/10.1037/arc0000032
(2014).
Das Wechselspiel von motorischer Leistungsfähigkeit und sportlichem Selbstkonzept [The interplay between motor performance and physical self-concept ]. In R. FrankI. NixdorfF. EhrlenspielA. GeipelA. MornellJ. Beckmann (Eds.), Performing under pressure (p. 102). Czwalina.(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6 (1), 1 – 55. https://doi.org/10.1080/10705519909540118
(2002). Changes in children’s self‐competence and values: Gender and domain differences across grades one through twelve. Child Development, 73 (2), 509 – 527. https://doi.org/10.1111/1467-8624.00421
(2015). Contrast and assimilation effects of dimensional comparisons in five subjects: An extension of the I/E model. Journal of Educational Psychology, 107 (4), 1086 – 1101. https://doi.org/10.1037/edu0000021
(2011). The importance of importance in the physical self: Support for the theoretically appealing but empirically elusive model of James. Journal of Personality, 79 (2), 303 – 334. https://doi.org/10.1111/j.1467-6494.2010.00678.x
(2021). An extension of the internal/external frame of reference model to the physical domain: Are there contrasting and assimilating achievement-self-concept relations in specific sports? Learning and Individual Differences, 90, 102048. https://doi.org/10.1016/j.lindif.2021.102048
(1986). Verbal and math self-concepts: An internal/external frame of reference model. American Educational Research Journal, 23 (1), 129 – 149. https://doi.org/10.3102/00028312023001129
(1990a). SDQ II manual: Self-description questionnaire – II. University of Western Sydney.
(1990b). Self Description Questionnaire I (SDQ I). Manual. University of Western Sydney.
(2007). Longitudinal study of preadolescent sport self‐concept and performance: Reciprocal effects and causal ordering. Child Development, 78 (6), 1640 – 1656. https://doi.org/10.1111/j.1467-8624.2007.01094.x
(2017).
Competence self-perceptions . In J. ElliotC. S. DweckD. S. Yeager (Eds.), Handbook of competence and motivation: Theory and application (pp. 85 – 115). The Guilford Press.(1994). A multidimensional physical self-concept and its relations to multiple components of physical fitness. Journal of Sport and Exercise Psychology, 16(1), 43 – 55. https://doi.org/10.1123/jsep.16.1.43
(2010). Maximum likelihood (ML) and diagonally weighted least squares (DWLS) estimation procedures: A comparison of estimation bias with ordinal and multivariate non-normal data. International Journal of Digital Society, 1 (1), 60 – 66. https://doi.org/10.20533/ijds.2040.2570.2010.0010
(2013). Dimensional comparison theory. Psychological Review, 120 (3), 544 – 560. https://doi.org/10.1037/a0032459
(2016). The generalized internal/external frame of reference model: An extension to dimensional comparison theory. Frontline Learning Research, 4 (2), 1 – 11. https://doi.org/10.14786/flr.v4i2.169
(2011). The reciprocal internal/external frame of reference model: An integration of models of relations between academic achievement and self-concept. American Educational Research Journal, 48 (6), 1315 – 1346. https://doi.org/10.3102/0002831211419649
(2020).
Selbstkonzept [Self-concept ]. In E. WildJ. Möller (Eds), Pädagogische Psychologie (pp. 187 – 209). Springer. https://doi.org/10.1007/978-3-662-61403-7_8(2006). Gender and course selection in upper secondary education: Effects of academic self-concept and intrinsic value. Educational Research and Evaluation, 12 (4), 323 – 345. https://doi.org/10.1080/13803610600765687
(2016). Global self-esteem, perceived athletic competence, and physical activity in children: A longitudinal cohort study. Psychology of Sport and Exercise, 22, 83 – 90. https://doi.org/10.1016/j.psychsport.2015.06.009
(2016). Measurement invariance conventions and reporting: The state of the art and future directions for psychological research. Developmental Review, 41, 71 – 90. https://doi.org/10.1016/j.dr.2016.06.004
(2020). Predicting global self-esteem in early adolescence: The importance of individual and gender-specific importance of perceived sports competence. Sport, Exercise, and Performance Psychology, 9 (4), 519 – 531. https://doi.org/10.1037/spy0000196
(2017). Music self-concept and self-esteem formation in adolescence: A comparison between individual and normative models of importance within a latent framework. Psychology of Music, 45 (6), 763 – 780. https://doi.org/10.1177/0305735616672317
(2015). Motor ability and self-esteem: The mediating role of physical self-concept and perceived social acceptance. Psychology of Sport and Exercise, 17, 15 – 23. https://doi.org/10.1016/j.psychsport.2014.11.006
(2008). Forschungsmethoden und Statistik in der Psychologie [
Research methods and statistics in psychology ]. Pearson.(1976). Self-concept: Validation of construct interpretations. Review of Educational Research, 46 (3), 407 – 441. https://doi.org/10.3102/00346543046003407
(2018). Examining chi-square test statistics under conditions of large model size and ordinal data. Structural Equation Modeling: A Multidisciplinary Journal, 25 (6), 924 – 945. https://doi.org/10.1080/10705511.2018.1449653
(1990). Math, verbal, and general academic self-concept: The internal/external frame of reference model and gender differences in self-concept structure. Journal of Educational Psychology, 82 (3), 546 – 554. https://doi.org/10.1037/0022-0663.82.3.546
(2015). Selbstwert und sportliches Fähigkeitskonzept im Jugendalter–der moderierende Effekt subjektiver Wichtigkeit. ZSF Zeitschrift für sportpädagogische Forschung, 3 (1), 61 – 83. https://doi.org/10.5771/2196-5218-2015-1
(1994). Exercise and self-esteem: Validity of model expansion and exercise associations. Journal of Sport and Exercise Psychology, 16 (1), 29 – 42. https://doi.org/10.1123/jsep.16.1.29
(1989). Exercise and self-esteem: Rationale and model. Medicine & Science in Sports & Exercise, 21 (3), 329 – 337. https://doi.org/10.1249/00005768
(2004). Die Messung des physischen Selbstkonzepts (PSK) [The Measurement of Physical Self-Concept (PSK) – The Development of the PSK-Scales for Children, Adolescents, and Young Adults]. Zeitschrift für Differentielle und Diagnostische Psychologie, 25 (4), 239 – 257. https://doi.org/10.1024/0170-1789.25.4.239
(2022).
Motorische Leistungsfähigkeit, Motorische Kompetenz, Selbstkonzept und physische Aktivität im Kindesalter [Motor performance, motor competence, self-concept, and physical activity in childhood. ]. In S. KlattB. Strauß (Eds.), Kognition und Motorik. Sportpsychologische Grundlagen und Anwendungen im Sport (pp. 226 – 240). Hogrefe.(2005). Zum Zusammenhang von Leistungen und Selbstkonzepten in verschiedenen Sportarten [Correlations between performance and perceived abilities in different sports]. Zeitschrift für Sportpsychologie, 12 (4), 135 – 143. https://doi.org/10.1026/1612-5010.12.4.135
(2008). Athletic classmates, physical self-concept, and free-time physical activity: A longitudinal study of frame of reference effects. Journal of Educational Psychology, 100 (4), 988 – 1001. https://doi.org/10.1037/0022-0663.100.4.988
(2006). Low self-esteem during adolescence predicts poor health, criminal behavior, and limited economic prospects during adulthood. Developmental Psychology, 42 (2), 381 – 390. https://doi.org/10.1037/0012-1649.42.2.381
(2018). The development of self-views across adolescence: Investigating self-descriptions with and without social comparison using a novel experimental paradigm. Cognitive Development, 48, 256 – 270. https://doi.org/10.1016/j.cogdev.2018.10.001
(2002). Toward a further understanding of and improvement in measurement invariance methods and procedures. Organizational Research Methods, 5 (2), 139 – 158. https://doi.org/10.1177/1094428102005002001
(2022). An individual participant data meta-analysis of the joint effects of social, dimensional, and temporal comparisons on students’ academic self-concepts. Educational Psychology Review, 34, 2569 – 2608. https://doi.org/10.1007/s10648-022-09686-1
(2019). The 2I/E model: Integrating temporal comparisons into the internal/external frame of reference model. Journal of Educational Psychology, 111 (7), 1131 – 1161. https://doi.org/10.1037/edu0000319
(2016). When men and women differ in self-esteem and when they don’t: A meta-analysis. Journal of Research in Personality, 64, 34 – 51. https://doi.org/10.1016/j.jrp.2016.07.007
