'It’s hard to define good writing, but i recognise it when i see it’: can consensus-based assessment evaluate the teaching of writing?

In a Higher Education environment where evidence-based practice and accountability are highly valued, most writing practitioners will be familiar with direct requests or less tangible pressures to demonstrate that their teaching has a positive impact on students’ writing skills. Although such evaluations are not devoid of risk and the need for them is contested, it can be argued that it is better to engage with them, as this can avoid the danger of overly simplistic forms of measurements being imposed. The current paper engages with this question by proposing the conceptual basis for a new measurement tool. Based on Amabile’s Consensual Assessment Technique (CAT), developed to assess creativity, the tool develops the idea of consensual assessment of writing as a methodology that can provide robust data through systematic measurement. At the same time, I argue consensual assessment reflects the evaluation of writing in real life situations more closely than many of the methodologies for writing assessment used in other contexts, primarily large scale tests. As such, it would allow writing practitioners to go beyond ethnographic methods, or self- reporting, in order to obtain greater insight into the ways in which their teaching helps change students’ actual writing, without sacrificing the complexity of writing as social interaction, which is fundamental to an academic literacies approach.