My thesis is situated in the field of language testing. Whenever an exam is reformed, changes in the classroom are likely to occur, that is, there may be a washback effect of external exams on what happens in the language classroom. Unlike washback on teaching or learning, washback on classroom assessment is under-researched. My study investigates the effect of a standardized Austrian school-leaving exam (the Matura), introduced in 2008, on classroom tests. The study thus follows several scholars’ call for research into how external tests affect assessment in the classroom (Wall and Alderson, 1993; Wall and Horak, 2006; Watanabe, 2000), and indeed addresses a research gap in our field.

The study specifically focuses on the washback of the exam’s reading paper on the assessment of reading in classroom-based tests. The research design for this thesis has two phases, each of which investigates the washback of the Matura on classroom tests from a different angle:

-         Phase 1 – comparison of characteristics of class test tasks

-         Phase 2 – teachers’ approaches for designing and selecting class test tasks

The study involves qualitative data collection methods (task analysis, interview) and quantitative analysis methods (statistical analyses) with the aim of increasing validity and reaching a more comprehensive understanding of examination washback on class test tasks. The data analysed not only includes self-report data (data about teachers’ approaches and views collected through interviews or questionnaires), but also includes the collection of observation data through an analysis of test tasks. This design meets the need for more objective measures for washback data.