Classroom dialogue: Does it really make a difference for student learning?

For at least two millennia, the form of dialogue that occurs in classrooms has been spotlighted as critical for teaching and learning. Moreover hypotheses have been proposed about optimal forms, with many scholars highlighting a need for open questions from teachers; extended responses from students (optimally involving justifications); cumulative building upon other students' contributions; articulation, discussion and evaluation of competing viewpoints; and gradual resolution of differences in a productive direction. However, despite its prominence in theoretical analyses, it remains uncertain whether classroom dialogue that takes this form is genuinely beneficial for student learning, and even if it is, whether the magnitude of benefit justifies making its promotion an educational priority. Although there is a handful of promising studies and findings are sufficiently encouraging to warrant further inquiry, past research has not been conducted with appropriate rigour nor on a large enough scale to provide conclusive answers.

Within the UK (as elsewhere), clarity about classroom dialogue is of the utmost importance, for dialogue is now an entrenched part of schooling: classroom talk is not limited to teacher monologue, and educators at all levels presume that this is appropriate. The implication is that, if the form of dialogue matters, it is already having significant but undocumented consequences. It is critical to ensure that these consequences are optimal. So recognizing the dearth of research and the urgent need for progress, this project will take a major step forwards. It is in fact the first study to investigate the implications of dialogue in whole-class settings. The project will use dialogue recorded in c.80 Year 6 classrooms. Three successive lessons will be videotaped in each classroom, with literacy, mathematics and science all covered. All teacher talk during target lessons will be recorded as will student talk that occurs during interaction with teachers. Small groups that are working independently of teachers will be sampled. To ensure a contribution to researcher development, postgraduate students (who will be trained and monitored) will assist with recording, and with transcribing and/or coding of data. Transcripts will be archived as a resource for future researchers.

Recordings will be coded and rated using a scheme that the applicants are developing. The scheme synthesizes the features with which scholars have characterized optimal dialogue, and while it goes beyond what any single analyst proposes, it is fully representative of the field as a whole. In other words, it is a model of effective practice that all analysts can sign up to, and it will be used in the project to represent degrees of approximation to a theorized ideal. It seems unlikely that any classrooms will reach the ideal. Nevertheless, published research together with the applicants' experience suggests that some will come close, while others will be nowhere near. The extent of approximation within the sample will be charted and specific areas of difficulty identified. Teachers' reflections will be invaluable, so workshops are planned where the 80 participants will be asked to brainstorm about why patterns of results were obtained, and what messages can be drawn for practice.

At the same time, analyses will assess whether there is a positive association between approximation to the theorized ideal and student learning, and if so whether the magnitude is sufficient to carry significant implications for policy. Taking steps to ensure that variation in classroom dialogue is not confounded through association with other predictors of learning, e.g. prior attainment/attitudes, SES, regression-style statistics will be used to assess the extent to which variation in dialogue predicts subsequent performance on objective, standardized and socially significant measures of learning attainment (e.g. Year 6 SATS) and attitudes to study.

Recent years have seen continuing and unresolved debate, publicly and internationally, about the best ways to teach. Much of that debate has focused on the extent to which learners should be encouraged to engage more actively and extensively in classroom talk than has traditionally, or typically, been the case in schools. Should students not merely be expected to attend to what teachers say but also to express their own ideas, ask questions and reveal any misunderstandings, so that these can be incorporated and dealt with through dialogue with the teacher (and taken into account in the teacher's plans for future lessons)? One reason the debate has been inconclusive is the lack of strong evidence to support the argument for reforming - or retaining - traditional teaching practices. By establishing whether or not a highly interactive, 'dialogic' approach helps to improve the quality of education and specifically learners' attitudes to study and attainment in curriculum subjects, the research will benefit teachers, policy makers, school inspectors and teacher educators the world over. It will allow educational policy and practice to be evidence-based to an extent that has not yet been possible and could inform media debates about teaching methods and the importance of teacher education.

The research will not only determine the effectiveness of dialogic methods, but also (potentially) allow the specification of teaching strategies that are associated with strong educational outcomes. These strategies could be exemplified through the project's observational data, which will consist of recordings that cover sequences of three linked lessons, allowing us to capture the temporal dimension of dialogue (Mercer, 2008). The strategies could be identified through using qualitative analysis of classroom talk in conjunction with the quantitative analysis that (as indicated elsewhere) lies at the heart of the project, to exemplify how effective teachers engage dialogically with their students in pursuit of particular topics or in seeking to overcome problems of understanding across one or more lessons in a sequence. The research would thus, in a precise way, inform the content of initial and continuing professional development for teachers. Work towards ensuring that envisaged benefits are realized will begin during the final three months of the project with workshops for teachers involved in the research. The teachers will be invited to brainstorm about why observed patterns of results were obtained, what the obstacles are and what support would be required should practice need to be developed. Such involvement will be crucial because, in our experience, teachers are often more receptive to the examples, recommendations and advice of expert teachers than to those of academic researchers. Outcomes will inform future workshops for other teachers.

More generally, the University of Cambridge is a very successful and high profile provider of initial teacher education (regularly rated 'Outstanding' by Ofsted), and by 2015 it will have set up a University Teaching School, which will develop and share exemplary practice. This will provide an excellent opportunity to showcase any effective pedagogic strategies which emerge from the project. The results may also inform the content of ITE and CPD programmes offered by other providers, e.g. through the regular and extensive presentations and workshops which members of the team make (see Pathways to Impact), and through proactive dissemination of outcomes to colleagues in other providing institutions. They will be made publicly available via the project website (and education blogs) and distributed via local networks such as the Cambridge Schools-University Research Partnership. The team members have successful experience of writing articles for non-academic audiences, and of contributing to broadcast media. The University's Press Office will be involved in planning how to optimize media coverage.