Trial lecture - time and place
Adjudication committee
- 1. opponent: Senior Researcher Mehmet Şükrü Bellibaş Leuphana University, Germany
- 2. opponent: Associate professor Stefan Johansson, Gothenburg University, Sweden
- Chair of committee: Professor Guri Skedsmo, University of Oslo, Norway
Chair of defence
Professor Ronny Scherer, Centre for Educational Measurement, University of Oslo, Norway
Supervisors
- Professor, Rolf Vegar Olsen, Centre for Educational Measurement, University of Oslo, Norway
Summary
Literature on school leadership is relatively scarce in terms of measurement and international comparisons, but fairly straightforward in terms of the importance of school leadership for the improvement of student learning. To properly measure and account for school leadership in a larger framework of schooling is therefore paramount. Through the lenses of an overarching theoretical, conceptual, and methodological framework, the thesis identifies five core challenges in measuring leadership in the international context, strongly reflecting on the Teaching and Learning International Survey (TALIS) design and data. The identified challenges referred to in this thesis vary from construct under-representations and the reliance on principals’ self-reported measures of school leadership, through the (in) comparability of leadership measures across populations and respondents, to the issues of whether we can quantitatively capture the heterogeneity of leadership in a reliable and valid way at the country level to answer whether leadership is a school or country level phenomena. Accordingly, this work suggests alternative approaches for improvement of the current practices by showing how the measure of school leadership can be constructed from a joint perspective of teachers, and how measurement properties of the construct of school leadership vary depending on whom within the school reports about them. The thesis also shows that leadership is more homogeneous in certain countries, but relatively universal in its various manifestations within countries internationally. Thus, inferences and comparisons at the level of countries are made and discussed considering the data limitations.
The empirical studies included in this thesis utilise data from teachers and principals participating in TALIS in the years 2013 and 2018, a quinquennial study administered by the Organization for Economic Cooperation and Development (OECD). The thesis consists of two parts, the extended abstract and three articles. The extended abstract discusses the background, purpose, and relevance of the thesis by presenting and discussing the relevant school leadership literature and literature on international large-scale assessments, in particular on TALIS. Therefore, the overarching research questions engage with the current conceptualization and measurement of school leadership in TALIS and future developments of the measurement of school leadership in the context of ILSA.
Article 1 investigates the measurement properties of the instructional leadership scale as used in TALIS 2013. Further, it explores the possibility to measure leadership from the teachers’ collective view by proposing two dimensions of leadership at the level of school, VI managing the instructional program and developing the school learning climate. As a robustness and validation check, the separate analyses are conducted in four Nordic countries. Article 2 expands the argumentation of Article 1 by examining how teachers’ and principals’ perceptions of school climate fit the framework of leadership for learning across 37 countries in TALIS 2018. Different perceptions of school climate are further investigated in relation to instructional and distributed leadership. Article 3 complements the findings from Article 2 by providing insights into leadership for learning at the system level. Clusters of schools with certain characteristics are summarised at the country level informing us about the relevance of systems for leadership research. With this approach, valid and relevant inferences about leadership are drawn in the context of countries. All three articles are based on advanced statistical analyses of TALIS data that are conducted in a multilevel setting but interpreted at the level of school. Overall, the findings presented in this thesis point to the current discussion of construct under-representation and the measurement invariance issues related to school leadership constructs in TALIS. The findings further suggest that different perceptions of school leadership within school matter for final measurement decisions.
This thesis informs those who work in the field of comparative leadership and international education that within-country leadership studies are more informative for actual leadership practice. Moreover, what matters greatly is from whom we collect data about leadership within the school encouraging a “more people involved” approach. The thesis also shows that leadership clearly is a school level phenomenon. However, the examination of the heterogeneity of leadership practices across countries, which reflects the broader cultural characteristics of societies, was not supported with the TALIS data. Therefore, we would rather talk about leadership as a global practice finely shaped in local schools and societies, than about leadership as a feature of educational systems and higher-level units. Finally, as a lateral finding, the thesis provides some more substantial knowledge on the associations between school leadership and other relevant factors at the school level.