Learning studies (LS) and Lesson studies have a common interest regarding development of teaching. Lesson study is basis also for LS, combined with ideas from design experiements (Marton 2014). Oneof the difference between the two is that in LS pre- and posttests are often used to get an input value of students’ knowing, as well as a learning outcome value after a certain intervention is performed. According to Pang and Marton (2003), one of the five steps of a LS is ascertaining students’ actual knowing, by an analysis of students’ conceptions or a pre-lesson test.Since these tests are emphasised as an essential ingredient in LS, the purpose of this paper is firstly to describe findings from a review of Swedish and Tanzanian LS (theses) in relation to the kinds of tests used and their purposes, and secondly to problematize such tests as method for exploring students learning.

Comparative analyses of the information about the pre- and post-tests in relation to the  learning object(s) in focus as well as of the purposes or functions of these tests were made in order to find patterns used for categorisation of types of pre and post-tests on the one hand, and functions of these on the other. Initial findings indicate that the pre and post-tests used vary from traditional paper and pen tests to semi/structured interviews, observations and video recorded lessons. Also the purposes for which they are used seem to vary. Furthermore, the language and praxis of testing seems to have been adopted in several studies without the necessary basis of test-construction. However, the idea of mapping students’ actual knowing as a of departure for specifying the object of learning, and for identifying possible critical aspects for learning and instruction is clearly motivated.  While some studies used learning studies in order to measure their effects,others have an interest in changing teaching in order to enhance learning. While the former types of LSneed to elaborate the relation between the learning object(s) andthe test(s) constructed (validity issues),as well as consider the statistical basis for their analyses, the latter types of LS could benefit from using a language better fit to these kinds of purposes.If the purposeis to become aware of students’ various ways of understanding the learning object and thereby the challenges for teaching then the issue is how to explore students’ knowing. In these cases LS can give substantial contributions to issues of formative assessment.

Marton, F. (2014). Necessary conditions of learning. London: Routledge