The International Reading Literacy Study PIRLS 2016 provides not only insights into the reading achievement of students towards the end of primary education, but also a view on reading instruction in grade 4 in class rooms in 50 countries. Good reading instruction is adapted to the needs of the students in the actual classrooms. It would therefore be misleading to try to search for causal relations between patterns of instructional activities and reading ability. Yet, it can be expected that activity patterns can be associated with specific class room contexts.

This presentation builds on the responses of approximately 250 teachers per country concerning the conditions and design of their reading instruction. Based on initial analysis for Austria (Schmich et al. 2017), an inventory of time use patterns and the didactics of reading instruction is built. Four core elements of reading instruction according to Garbe (2016) were examined (1): promotion of reading fluency by reading aloud, (2) teaching of reading strategies, (3) promotion of higher reading quantity and sustained reading, (4) reading animation. Parallel class rooms are classified by their composition of reading ability. Classrooms with homogenous and heterogeneous abilities at different levels of reading literacy were identified, searching for patterns of instruction related to specific reading contexts. The analysis is done at a country by country level and aims to identify country specific uses of instructional approaches. It addresses the question of whether (a) reading instruction in high-performing classes differ from teaching in low-performing classes, and (b) reading-related activities in the classroom vary between performance-homogenous and performance-heterogeneous classes.

In a first step, teacher data on reading instruction are used to construct measures reflecting the four core elements of reading instruction. Using latent class analysis, common patterns of instruction are distinguished and described. The latent class analysis is done for a pooled data set reflecting the width of countries participating in PIRLS. The analysis is then validated by exemplary country wise construction of latent classes. Secondly, the class rooms are categorized according to the reading competence of the pupils and characterized in terms of their background features. Finally, class rooms are cross classified with respect to the latent classes of reading instruction methods and the ability composition of the students. This leads to a landscape of instruction pattern for participating countries.

It is assumed that different teaching activities take place within the contexts of differentiated reading lessons in high-performing and low-performing classes; e.g., more frequent teaching of decoding strategies in low-performing classes and more frequent reading in longer literary texts in high-performing classes. Also, differences in reading instruction between performance-homogeneous and performance-heterogeneous classes are expected, e.g. with regards to the use of differentiated reading texts. Understanding common patterns of instruction for different class compositions can contribute to discussions with practitioners on the use of assessment results for teaching. An international comparative perspective on the identified pattern will help to reconfirm or question national policies.

Schmich, J., Breit, S., Lanzdorf, R., & Itzlinger-Bruneforth, U. (2017). Schulischer Kontext: Leseunterricht, Fortbildung und Schulressourcen. In: C. Wallner-Paschon, U. Itzlinger-Bruneforth & C. Schreiner (Eds.). PIRLS 2016. Erste Ergebnisse. die Lesekompetenz am Ende der Volksschule (p. 83-101). Graz: Leykam. Garbe, C. (2016). Didaktische Aspekte des Kompetenzbereichs „Lesen – Umgang mit Texten und Medien“. In BIFIE (Eds.), Themenheft für den Kompetenzbereich „Lesen – Um- gang mit Texten und Medien“ (p. 5-20). Graz: Leykam.