This presentation focuses on the use of computerized dynamic assessment as a means of diagnosing learners' responsiveness to pedagogical intervention. We discuss two recent studies of Chinese pragmatic comprehension to illustrate how different administration formats highlight different aspects of responsiveness to instruction.
A defining feature of diagnostic assessment is its focus on identifying learners’ strengths and weaknesses. This information can be used to inform future pedagogical interventions. Dynamic assessment aims to uncover what learners can do alone but also what becomes possible with mediation (i.e., instructional support), with an eye toward understanding how learners may respond to instruction. This paper addresses the use of computerized dynamic assessment (C-DA) as a diagnostic tool for understanding responsiveness to instruction as evidenced by learner improvements over the course of the assessment exercise. Drawing on two recent studies of second language (L2) Chinese pragmatic comprehension that focused on comprehension of indirect speech acts (e.g., understanding that the utterance “It’s warm in here” can be an indirect request to turn down the heat), we illustrate two approaches to diagnosing responsiveness to instruction in C-DA contexts. In the first example, mediation was provided in a so-called ‘cake format’ (Sternberg & Grigorenko, 2002; Lantolf & Poehner, 2004), meaning that instructional support was available throughout the assessment. Such a format allowed us to track responsiveness to instruction in real time as learners take the test (e.g., does a learner improve from item to item?). In the second example, mediation was sandwiched between a pretest and a posttest. This enabled us to draw a clearer distinction between learners’ abilities at the start of the assessment, and what became possible after mediation. Following an overview of the methods and results of our work to date, we discuss the advantages and disadvantages of the two test administrations along two dimensions: 1) practical/logistical concerns (e.g., complexity of designing tests) and 2) diagnostic/pedagogical concerns (e.g., [in]dependence of assisted and unassisted performance scores). We offer a number of recommendations for using each approach to C-DA for diagnostic purposes.
