Teachers’ instruction is at the heart of education, and previous research has shown that teaching quality is important for students’ learning outcomes (e.g. Charalambous & Praetorius, 2020; Seidel & Shavelson, 2007). However, teaching is a two-way process, and less is known about how the composition of the classroom affects teaching quality (TQ). Do for instance high socio-economic (SES) classrooms receive different TQ than low-SES classrooms? To examine this, one would first need to establish whether a so-called compositional effect exists. Compositional effect refers to the effects of, for instance the classroom’s socio-economic status (SES) on student learning outcomes, over and above the effect of students’ individual SES (Van Ewijk & Sleegers, 2010).

Both compositional effects and unfair distribution of high-quality teachers have been found in previous studies in a number of countries (Gustafsson et al., 2018; Luschei & Jeong, 2018; Van Ewijk & Sleegers, 2010) However, in Norway, that for a long time was considered an egalitarian society (Buchholtz et al., 2020), there is a lack of such studies. At the same time, educational inequality has increased in Norway (Sandsør et al., 2021). Hence, the overarching aim of the present study is to examine whether a compositional effect exists, and how the composition of the classrooms affects TQ. We further aim to describe more in depth what characterizes the TQ in classrooms of different compositions in Oslo where the gaps between students are larger and there are more minority students than in the rest of Norway (Fløtten et al., 2023).

The following research questions were asked:

1)      What is the effect of the classroom composition (in terms of SES and minority status) on students learning outcomes in science, over and above students’ individual SES and minority status (i.e. the compositional effect)? How does this differ between Oslo and the rest of Norway?

2)      What is the effect of the classroom composition on TQ in science, and how does this differ between Oslo and the rest of Norway?

3)      What characterizes TQ in science classrooms of different compositions in Oslo?

Theoretical framework for teaching quality.

We chose The Three Basic Dimensions (TBD) framework (Klieme et al., 2009; Praetorius et al., 2018) to conceptualize TQ as this framework is the most commonly used in Europe and by the international large-scale studies (Klieme & Nilsen, 2022). TQ is here defined as the type of instruction that predicts students learning outcomes, and includes the following three dimensions:

1) Classroom management refers to how teachers manage the classroom environment and includes, for instance, preventing undesirable behaviors and setting clear and consistent rules and expectations for student behavior.

2) Supportive teaching focuses on the teacher’s ability to support students both professionally and socio-emotionally, such as providing clear and comprehensive instruction and seeing and listening to every individual student.

3) Cognitive activation includes instruction that enables students to engage in higher-level cognitive thinking that promotes conceptual understanding. Such instruction is characterized by challenging and interactive learning.

The TBD is a generic framework used across subject domains. To address research question 3, and further investigate more in depth the subject-specific aspect of TQ in science, a fourth dimension from the framework Teacher Education and Development Study–Instruct (TEDS-Instruct, e.g. Schlesinger et al., 2018) was included. This framework was adapted to the Norwegian context and to the subject domain of science, and validated. The fourth dimension is called Educational structuring and refers to subject-specific aspects of instruction such as inquiry or dealing with students misconceptions in science.

Design and sample. The project Teachers’ effect of student learning (TESO), funded by the Norwegian Research Council, collected data through an extended version of TIMSS 2019, including a representative sample of fifth graders in Norway, a representative sub-sample of Oslo, and video observations of grade six classrooms in Oslo. The students who participated in the video observations in sixth grade, also participated in TIMSS 2019 when they were fifth graders. All students answered questioners and the TIMSS mathematics and science tests. Measures. To measure the generic TQ in the second research questions, students’ responses to the questionnaire were used. In the questionnaire, Classroom management was measured by 6 items (e.g. “Students don’t listen to what the teacher says”). Cognitive activation was measured by 5 items (e.g. “The teacher asks us to contribute in planning experiments”. Both of these were measured using a 4-point frequency scale (from Never to Every or almost every lesson). Teacher support included 6 items on 4-point Likert scales (from Disagree a lot, to Agree a lot), e.g. “My teacher has clear answers to my questions”. To answer research question 3 and provide more in-depth descriptions of TQ, the more fine-grained TEDS-Instruct observation manual (including 21 items ratted from 1 through 4) was used to rate the videos. The manual measures the same three aspects as TIMSS conceptually, in addition to educational structuring. SES was measured by students’ responses to the number of books at home (the parents’ responses to their education had more than 40% missing and was hence excluded as a SES indicator). Minority status was measured by students’ answer to how often they speak Norwegian at home. Methods of analyses To answer research questions 1 and 2, we employed multilevel (students and classes) structural equation modelling (SEM), and a multi-group approach to examine differences between Oslo and the rest of Norway. To avoid multi-collinearity, each aspect of teaching quality was modelled separately and as latent variables. Compositional effects were estimated by subtracting the within level effects from the between level effects. To answer research question 3, the questionnaires, achievements, and ratings on the videos were linked and merged to one file. Descriptives were used to crate profiles of the ratings of the video observations to describe the characteristics of TQ in classrooms of different compositions.

RQ1. Compositional effects The compositional effects were all significant (p< .05) and positive. The effect of SES was 0.44 for Norway, and the multigroup analyses yielded an effect of 0.57 for Oslo and 0.31 for the rest of Norway. The compositional effects of language were 0.45 for Norway, 0.76 for Oslo and 0.45 for the rest of Norway. In other words, the compositional effects for Oslo were very high, while the compositional effects for Norway overall were in line with other Scandinavian countries (Yang Hansen et al., 2022). RQ2. Relations between classroom composition and TQ High-SES, and especially low minority classrooms, had positive and significant associations to both classroom management and teacher support. These effects were stronger in Oslo than the rest of Norway. This indicates an unfair distribution of high teaching quality to advantaged classrooms. However, for cognitive activation, there were no significant results at the class level, but a negative association between high-SES, low-minority classrooms and students’ perceptions of cognitive activation. This indicates that advantaged students perceive less challenge and interactive learning. RQ3. Characteristics of TQ Results from the video observations showed that TQ in high-SES classrooms were characterized by better classroom management, teacher support, and educational structuring than low-SES classroom, albeit with less cognitive activation. Furthermore, high SES classrooms were characterized by fewer minority students and higher achievements than low SES classrooms. These findings are in line with the results from the questionnaires. Taken together, the findings from our three research questions points to a school that contributes to increase the gap between students. Classrooms with high shares of advantaged students have access to better teaching quality than classrooms with many disadvantaged students, thus generating unequal opportunities to learn.

Buchholtz, N., Stuart, A., & Frønes, T. S. (2020). Equity, equality and diversity—Putting educational justice in the Nordic model to a test. Equity, equality and diversity in the Nordic model of education, 13-41. Charalambous, C. Y., & Praetorius, A.-K. (2020). Creating a forum for researching teaching and its quality more synergistically. Studies in Educational Evaluation, 67, 100894. Fløtten, T., Kavli, H., & Bråten, B. (2023). Oslo er fortsatt en delt by [Oslo is still a divided city]. Aftenposten. Retrieved from https://www.aftenposten.no/meninger/kronikk/i/dw2z8o/oslo-er-fortsatt-en-delt-by Gustafsson, J.-E., Nilsen, T., & Hansen, K. Y. (2018). School characteristics moderating the relation between student socio-economic status and mathematics achievement in grade 8. Evidence from 50 countries in TIMSS 2011. Studies in Educational Evaluation, 57, 16-30. Klieme, E., & Nilsen, T. (2022). Teaching Quality and Student Outcomes in TIMSS and PISA. International Handbook of Comparative Large-Scale Studies in Education: Perspectives, Methods and Findings, 1089-1134. Klieme, E., Pauli, C., & Reusser, K. (2009). The pythagoras study: Investigating effects of teaching and learning in Swiss and German mathematics classrooms. The power of video studies in investigating teaching and learning in the classroom, 137-160. Luschei, T. F., & Jeong, D. W. (2018). Is teacher sorting a global phenomenon? Cross-national evidence on the nature and correlates of teacher quality opportunity gaps. Educational researcher, 47(9), 556-576. Praetorius, A.-K., Klieme, E., Herbert, B., & Pinger, P. (2018). Generic dimensions of teaching quality: The German framework of three basic dimensions. ZDM, 50(3), 407-426. Sandsør, A. M. J., Zachrisson, H. D., Karoly, L. A., & Dearing, E. (2021). Achievement Gaps by Parental Income and Education Using Population-Level Data from Norway. https://osf.io/preprints/edarxiv/unvcy Schlesinger, L., Jentsch, A., Kaiser, G., König, J., & Blömeke, S. (2018). Subject-specific characteristics of instructional quality in mathematics education. ZDM, 50, 475-490. Seidel, T., & Shavelson, R. J. (2007). Teaching effectiveness research in the past decade: The role of theory and research design in disentangling meta-analysis results. Review of Educational Research, 77(4), 454-499. Van Ewijk, R., & Sleegers, P. (2010). The effect of peer socioeconomic status on student achievement: A meta-analysis. Educational Research Review, 5(2), 134-150. Yang Hansen, K., Radišić, J., Ding, Y., & Liu, X. (2022). Contextual effects on students’ achievement and academic self-concept in the Nordic and Chinese educational systems. Large-scale Assessments in Education, 10(1), 16.