The effect of class composition by gender and ability on secondary school students’ school well-being and academic self-concept: A literature review

Abstract

In the field of educational effectiveness research, the influence of a class’ student body on students’ individual achievement scores has been a popular research interest for many years. Yet, few studies have focussed on the effects of class composition on students’ non-achievement outcomes, and up to now, hardly any attempts have been made to summarize the findings of previous studies on this topic by conducting a literature review. The current study tries to fill in this gap and focuses on the effects of class composition in secondary education in terms of ability and gender on students’ school well-being and academic self-concept. The results of this literature review indicate that ability grouping is beneficial for strong students’ school well-being, but rather detrimental for the school well-being of weak students. The reverse holds for students’ academic self-concept. Furthermore, our results show that single-sex classes are advantageous for girls’ school well-being and academic self-concept. As for boys, the results are inconclusive.

Introduction

Since the influential Coleman report (Coleman et al., 1966) first brought the topic into the spotlight, the effect of classmate background characteristics (such as ability and gender) on the achievement of individual students has been thoroughly discussed in educational effectiveness research (Hattie, 2002). However, it is only recently that non-achievement outcomes are considered important as well, and viewed as educational objectives in themselves, since students with low levels of school well-being and academic self-concept are at risk for depressive reactions, lower grades, school dropout, and delinquency (Van Houtte, 2005). A noticeable amount of research has not only concentrated on achievement effects but also on the non-achievement effects of schools and classes. Although the results of these studies are somewhat conflicting about the consistency between achievement and non-achievement outcomes at school level, there exists a certain amount of proof for the fact that classes effective in enhancing student achievement are not necessarily effective in enhancing the school well-being and academic self-concept of its students (Hofman et al., 1999, Opdenakker and Van Damme, 2000). Furthermore, it has been demonstrated that classes have a larger effect on students’ achievement than on non-achievement outcomes. For example, Opdenakker and Van Damme (2000) established that 23% of the variation in students’ language achievement at the end of the first grade of secondary school could be attributed to the class-level, while only 4% and 2% variation occurred, respectively, in students’ school well-being and academic self-concept between classes within schools. Although these percentages are quite small, they do show that classes have an effect on students’ non-achievement outcomes and can make a difference.

An important factor that differentiates one secondary class group from another is its student composition in terms of ability and gender. The objective of the present study is to examine the effects of the composition of a secondary school’s class group in terms of these two features on students’ school well-being and academic self-concept, by comparing the findings of previous research. We define our two independent and two dependent variables as follows.

Class composition by ability refers to the practice of purposely grouping students into classes on the basis of their abilities. A distinction can be made between within- and between-class ability grouping. In this study we focus on between-class ability grouping, commonly referred to as tracking or streaming. This practice occurs in secondary schools in many countries, but the nature and extent of between-class ability grouping differs across countries, making the concept somewhat ambiguous. Prior achievement is frequently the most important determining factor in class composition practices by ability, but measures of general ability, academic specializations, and teacher judgements may also play a role. Thus, ability-grouped classes are typically distinguished by average achievement levels and/or a specific curriculum. Furthermore, between-class ability grouping can be used for (1) just a few subjects also referred to as setting, implicit tracking or opt-in tracking or (2) all subjects, also called explicit tracking. Whereas setting is common in countries such as the United States, the United Kingdom, Canada, and Australia, explicit tracking is more widespread in countries like Japan, Belgium, the Netherlands, and Germany. In countries in which setting is used, the impact of ability grouping on later educational outcomes is less visible and rigid, although secondary school students’ choices of advanced placement courses are associated with their future educational career (Trautwein, Lüdtke, Köller, Marsh, & Baumert, 2006). Furthermore, in setting endorsing countries ability grouping is generally not prescribed by law. By contrast, in countries in which explicit tracking is common, the tracking status of a student is clearly visible to the student, his or her peers, parents, and teachers and has a profound impact on later educational opportunities (Trautwein et al., 2006). In addition, in these countries, ability grouping is usually prescribed by law.

This background information should be kept in mind when interpreting the results presented in this study as it is well imaginable that the explicitness of a student’s tracking status has an impact on the way a student perceives himself/herself and thereby on his/her school well-being and academic self-concept. This could explain possible contradictions in research findings stemming from different countries.

One important motive for grouping students by ability is the assumption that teaching a group of students that is fairly homogeneous in ability is more efficient than catering to the diverse needs of a heterogeneous ability group. As such, tracked education would be more beneficial for all students’ achievement development (Hallinan, 1994). However, research on the effect of tracking on student achievement has shown the opposite. Several studies have found that belonging to a high ability class positively influences students’ academic achievement, whereas the opposite holds true for belonging to a low ability class (Hallam and Ireson, 2003, Hallinan, 1994, Van de gaer et al., 2006). This negative effect of tracking on the achievement of low-ability grouped students might be (partially) explained by students’ non-achievement outcomes. It is very well imaginable that being placed in a lower track goes paired with feelings of status loss, interest loss, and lower well-being, and that this, in turn, might negatively affect students’ academic achievement. In our review, we will discuss how ability grouping influences students’ non-achievement outcomes.

Over the past few decades, a number of differences in the achievement scores and educational aspirations of boys versus girls have come into light. Firstly, girls outperform boys on almost every school subject (Van de gaer, Pustjens, Van Damme, & De Munter, 2004). Secondly, girls are underrepresented in science, technology, engineering, and mathematics (together referred to as STEM; Kessels & Hannover, 2008). Research has suggested that schools contribute to these gender differences, both through the content of the curriculum and the way it is presented, as well as through gendered interactions between students (Kelly, 1985). Among a number of strategies that have been implemented in order to combat this gender gap in academic achievement and aspirations, is the realization of single-sex classes within coeducational schools and entirely single-sex schools. The rationale behind this homogeneous grouping by gender is that teachers can adapt their teaching methods and subject material to the different learning styles and preferences of boys versus girls. By these adaptations, students’ learning enjoyment should be enhanced. Furthermore, single-sex classes are considered to offer a better learning climate by providing an academic environment that makes it more possible for students to concentrate on learning tasks because there are less non-academic distractions and disciplinary problems. Moreover, in single-sex classes, girls are believed to be less discriminated (Hattie, 2002). In our review, we will examine whether the learning adaptations that are made in single-sex classes indeed foster students’ school well-being and academic self-concept. Two types of studies on gender grouping were included: (1) studies comparing single-sex classes with coeducational classes within coeducational schools (e.g., all-girls classes for mathematics) and (2) studies comparing single-sex schools (with inevitably single-sex classes) with coeducational schools (always with coeducational classes). So, in fact, this last type of studies also compared single-sex classes with coeducational classes. The assumption here is that being part of a single-sex class has the same effect on school well-being and academic self-concept regardless of whether the class is situated in a single-sex or coeducational school. However, we treated this assumption with utmost care, and prudently compared the results of studies of type (1) with the results of studies of type (2).

Various definitions of school well-being exist. Hofman et al. (1999) consider well-being at school as the attitude of students towards the attended school. This attitude towards school can be divided into four dimensions: a general attitude towards school-life, attitudes towards teachers, attitudes towards peers, and attitudes towards the school organization and building. Samdal, Wold, and Bronis (1999) suggest that in addition to these four attitude aspects, school well-being also includes the degree to which a student feels safe and secure at school (in terms of not feeling lonely and not being bullied), and the degree to which students feel that teacher expectations are reasonable. Eder (1995) likewise acknowledges the degree to which a student feels safe at school as an indicator of school well-being. He refers to school well-being as the degree to which a student feels good at school, is satisfied with aspects of the situation he/she is in at school, and is free of school-related feelings of fear and psychological or psychosomatic problems induced by the school situation.

Engels, Aelterman, Van Petegem, and Schepens (2004) conceive students’ perceptions of the school atmosphere, the school infrastructure and students relationships with their teachers as indicators of school-wellbeing in addition to the general connotation of liking school.

Although these definitions of school well-being differ to a certain extent, they all emphasize the student’s affective attitude toward school as a place of learning and living. Therefore, in this literature review, the term school well-being is broadly defined as the degree to which a student likes school. Studies that focused on one of the aspects of school well-being as mentioned in the previously described definitions were included in this literature review.

School well-being is generally measured by questionnaires. For example, Engels et al. (2004) developed a school well-being questionnaire for Flemish secondary school students, including items referring to the general connotation of liking school such as: “I usually like going to school”, “I generally feel good at school” and “I really like my school” as well as questions referring to specific aspects of school well-being such as “Are you satisfied with the atmosphere at your school?”, “Are you satisfied with your school building?”, and “Do the teachers in your school treat you with respect”?

Academic self-concept refers to students’ knowledge and perceptions about themselves in achievement situations (Bong & Skaalvik, 2003). A student’s academic self-concept relies largely on social comparative information and reflected appraisals from significant others and is usually measured by questionnaires with items such as “Compared with others my age, I’m good at (a subject)” or “In (a subject), I am one of the best students in my class” (e.g., Marsh, 1999).

Academic self-concept is a multidimensional construct, meaning that self-concepts in specific domains (e.g., mathematics) are subsumed under self-concepts in more general domains (e.g., academic self-concept) (Bong & Skaalvik, 2003).

Different views exist among researchers whether academic self-concept also encloses emotional reactions to learning tasks such as interest, enjoyment, and satisfaction. Some consider these as part of academic self-concept (Marsh, 1999), whereas others consider them as distinct constructs (Bong & Skaalvik, 2003). In this review, therefore studies differed to some extent in their conceptualization of academic self-concept.

In order to summarize the research findings on the impact of secondary school class composition on the two non-achievement outcomes, six research questions are addressed:

• Does class composition in terms of ability have an effect on secondary school students’ school well-being and academic self-concept?

• Does class composition in terms of ability have differential effects for high-ability and low-ability secondary school students?

• Does class-composition in terms of ability have differential effects in the case of explicit tracking versus of implicit tracking?

• Does class composition in terms of gender have an effect on secondary school students’ school well-being and academic self-concept?

• Does class composition in terms of gender have differential effects for male and female secondary school students?

• How can possible class composition effects be explained?

The remainder of this article consists of three sections. Section 2 describes how the studies included in this literature review were identified and selected. Section 3 presents the empirical findings concerning the effects of classes with a dominant student population in terms of ability and gender on the previously mentioned two student non-achievement outcomes. In addition, possible explanations for these class composition effects are mentioned. Section 4 summarizes the main findings and addresses practical implications and avenues for future research.