Background
Achievement differences across schools have increased in Sweden during the past two decades, at the same time as average achievement levels have declined, for example in mathematics (Holmlund et al., 2014; SOU 2014:05) and reading (Skolverket, 2013). These negative trends may be related to schools’ demographic, organizational and resource prerequisites. The organizational basis for the Swedish compulsory school has undergone significant changes since the late 1980s, with free school choice, an increasing number of private schools and a larger autonomy for school leaders. The increased school segregation with respect to migration background since the implementation of these reforms (Yang Hansen & Gustafsson, 2016), could possibly be caused by the continuing decline in inclusion of migrant students and a related educational inequality in instructional quality and teacher competence. Inclusion is defined as a structure of organizing integration according to particular rules and regulations.
The results of research on effects of teacher competence are, however, somewhat fragmented and unclear, at least with respect to effects on different student groups. This can partly be explained by uncertainties in the determination of crucial teacher characteristics. Making comparisons between for example authorized and unauthorized teachers has been shown to be problematic. There is vast research in this field which is characterized by severe methodological problems, e.g. with respect to drawing causal conclusions from cross-sectional observational data. For example, no correlation between student achievement and various resource factors has been found in some studies (e.g. Hanushek, 1997), whereas others found a positive correlation (e.g. Greenwald, Hedges & Laine, 1996).
A teacher effect on student achievement is, however, well manifested (e.g., Gustafsson, 2003; Gustafsson & Myrberg, 2002; Johansson, Myrberg, & Rosén, 2015; Nye, Konstantopoulus, & Hedges, 2004; Rockoff, 2004), and the results also suggest that lower achieving students, as for example immigrants, are the more likely to benefit from increases in teacher effectiveness (e.g., Sanders, 1998). Teacher quality is, furthermore, one of the resource factors that explains most of the increase in performance differences between schools in Sweden (Björklund, Fredriksson, Gustafsson, & Öckert, 2010, Ch 7; Gustafsson & Myrberg, 2002). The general reduction in teacher quality in Sweden the last decades (SOU 2014:05), and the decreased equality of allocation of teacher competence between schools (Hansson & Gustafsson, 2016; OECD, 2013), supports these results. However, the variation between student outcomes that different teachers are achieving (Hanushek, 2003) needs to be further problematized and discussed. Teacher knowledge and skills, teacher training and teaching experience are examples of characteristics highlighted in different meta-studies (e.g., Greenwald, Hedges & Laine, 1996), that should be subject of such an investigation.
This study intends to develop a precise and differentiated description of teacher quality for use in future analyses of relations between teacher competence and educational results, with focus on interactions with student composition of schools due to students’ socio-economic and migration backgrounds. The description is among other factors focusing on teacher’s basic knowledge, subject-related and pedagogical training, and type of teacher training program. One aim is to investigate the variation over time in access to qualified teachers and the variation in teacher qualifications between schools. In further analyses, the significance of the teacher characteristics for literacy and mathematics in grades 1 to 6 will be focused.
Overall, the project, which this study is a part of, is expected to generate insights about essential conditions for effective and equitable teaching in Swedish and mathematics in primary school, and about distributions of teacher competence across schools with different student composition.