Comparing summative and dynamic assessments of L2 written argumentative discourse: Microgenetic validity evidence

Abstract

Drawing upon multiple data sources, this case study compares the summative assessment (SA) and dynamic assessment (DA) of L2 writing, in order to examine the validity of DA. There is a dearth of research on the validity of DA of L2 writing. Most DA studies to date have examined its effectiveness in assessing and promoting learning based on a loosely defined construct, which DA aims to assess and develop, and without discussing DA’s validity in relation to the microgenesis that is expected to transpire within DA. This study examines what an L2 learner knows of the constructs and can do with mediation to produce argumentative texts. The two constructs which were assessed and promoted were an adaptation of the Toulmin (1958/2003) model of argument and the knowledge-transforming cognitive composing process (Bereiter & Scardamalia, 1987) that led the production of the texts. After a period of instruction in argumentation and its cognitive processes, SA suggested that the learner’s development, as judged by his scores, was almost flat; thus, DA was performed to verify if, indeed, there was no true development and to simultaneously assess if, with mediation, the learner could produce the missing argumentative discourse features; a series of dialogic and contingent DA interactions were enacted to help the learner identify and produce the missing features; both goals were successfully accomplished. A microgenetic, together with rating and textual analysis, was employed to examine the quality and validity of mediation required to generate the discourse features. An argument-based approach to validity supports the claim that DA complemented SA to make the assessment practice fairer by tapping into emergent learning; the argument approach provides a robust basis for the interpretation and use of the findings of DA.