A Framework for Assessment for Learning: Implications for Feedback Practices within and beyond the Gap

Three recurring emphases in the literature on formative assessment are (a) the importance of assessment design in prompting and sustaining students' learning, (b) giving students feedback that enables them to improve their work, and (c) clarity of standards to articulate the gap between past and desired performances. Much has been written on how each of these is important in designing and using assessment for learning. But there is considerably less attention (if any) on how these emphases should confluence each other. In this paper, I propose a framework for assessment for learning (AfL) wherein assessment standards (vertical axis), task design (horizontal axis) and feedback practices (incline) form a “triangle of practices” that construct the “space” for enhancing learning. Implications for formative assessment to extend learning beyond its triangulated frame are discussed. 1. Introduction Black and Wiliam’s [1] comprehensive review of formative assessment research has been relied on by many to argue that formative assessment has significant benefits for enhancing students’ learning. In that study, the authors framed formative assessment (very) broadly to be interpreted as “encompassing all those activities undertaken by teachers, and/or by their students, which provide information to be used as feedback to modify the teaching and learning activities in which they are engaged.” (Page 7). Such a broad notion of formative assessment included diverse practices such as self-assessment, learning goals, and mastery learning. This casts doubts on whether it is possible to precisely determine the specific “formative assessment practice” for which the gains in performance could have been attributed to. Bennett [2] argues forcefully: a major concern with the original Black and Wiliam [1] review is that the research covered is too disparate to be summarised meaningfully through meta-analysis. That research includes studies related to feedback, student goal orientation, self-perception, peer assessment, self-assessment, teacher choice of assessment task, teacher questioning behaviour, teacher use of tests, and mastery learning systems. That collection is simply too diverse to be sensibly combined and summarised by a single, mean effect-size statistic (or range of mean statistics). (Page 11). The ambiguity over what is exactly and minimally needed for formative assessment to succeed is not confined to Black and Williams [1]. Shepard [3] produced a comprehensive model of formative assessment which lists ten key characteristics of formative assessment of which Cizek