Cooperative learning has been found to be effective in many studies, so how would it perform in England? The results were surprising, as Robert Slavin, Mary Sheard, and Pam Hanley explain
One of the most reliably effective approaches to the teaching of mathematics is cooperative learning. In methods that have been found to be effective, students work in 4-5 member groups to help one another master mathematical concepts and operations, following teacher teaching. Students are then individually assessed based on their learning of the content, and teams receive recognition based on the average scores of all team members. Since the late 1970s, study after study has found methods of this type to accelerate mathematics learning in elementary and secondary schools in the US.
Based on this experience, we set out to create and evaluate a cooperative learning approach to mathematics that would take cooperative learning to the next level and evaluate it in England, where relatively little quantitative evaluation of cooperative learning had taken place. It was an ideal project for the newly opened Institute for Effective Education (IEE) at the University of York. The IEE was established to use rigorous experiments to evaluate programs for use in the UK, and to disseminate information on effective practices to a UK audience. The IEE carried out a large-scale evaluation of Student Teams-Achievement Divisions, or STAD, a simple cooperative learning approach. Part of the intent of the study was just to show whether cluster randomized experiments – in which schools are randomly assigned to treatments – could be done in England, which lacked much of a tradition of quantitative experiments. We evaluated cooperative learning because we were pretty sure it would work.
The experiment worked – in the sense that we were able to carry out the evaluation. However, the outcomes were disappointing. Pupils in Years 4 (age 8-9) and 5 (age 9-10), who learned by working in cooperative groups, gained no more than controls when measured using Optional SATs (tests aligned with the Year 6 (age 10-11) SATs that are the main accountability measure in English primary schools).
This result was most surprising, and we were determined to try again. With funding from the Nuffield Foundation and the Bowland Charitable Trust, we developed and evaluated a new cooperative learning model specifically for the English context. The approach was called PowerTeaching Maths (PTM). In addition to the use of cooperative learning, PTM provided electronic “flip charts” for teachers to use on interactive whiteboards, which are almost universally available in England. We made appealing video vignettes to introduce each topic, with characters such as Shirleylock Holmes, Tip Topper, and Grandpa and Dragon setting up practical contexts for solving math problems. Our talented developers and trainers piloted the program in schools in two regions of England for a year, and then revised it in light of what they saw and what teachers told us.
Finally, we set up a large cluster randomized experiment, with 42 schools assigned at random to use PTM or to continue with their usual teaching methods. Children were pre- and post-tested on Optional SATs. In addition, observers watched teachers and recorded the degree to which they implemented the program.
When the statistical analyses came in we were terribly disappointed. Once again, pupils in the cooperative learning condition (PTM) gained no more than those in the control group. Yet, in general, we and our observers had seen happy, productive children, teachers were quite positive about PTM, and quality of implementation overall was rated as adequate. What happened?
As we looked more closely at teachers’ feedback and observations, two big issues jumped out. Both go to the core of differences between traditional teaching practices in England and in North America, where cooperative learning began.
One issue related to far greater reluctance among teachers in England to do formal assessments of pupils’ learning. Brief assessments, or quizzes, are the final step in every PTM (or STAD) lesson, and US teachers are very comfortable with regular assessments. The quiz scores in cooperative learning are used to give feedback to the teams; the average scores on these individual quizzes contribute to team scores, and the quiz process is what focuses pupils on trying to ensure that every member of their team has mastered the content.
Our observations found that teachers in England usually skipped this essential step. Instead, they gave teams points based on pupils’ behaviour and helpfulness, but not on the actual learning of all team members.
The second issue related to differentiation. In cooperative learning, teams are composed of students at all levels of performance, and teachers teach the same lesson to all. In England, this is rarely done. Teachers are expected to prepare variations of their lessons, for children below, at, or above the expected norms. The teachers in the study found it very difficult to teach one lesson to all, and many complained that PTM was holding back high achievers and (less often) failing to attend adequately to low achievers.
The findings of our studies of cooperative learning in English primary schools were not what we had hoped for, but they taught us an important lesson. Teaching methods proven to be effective in one culture and system cannot be assumed to be effective in another. We remain convinced that the cooperative learning strategies that have been found in North American research to be effective in mathematics can be made to work in England, but they are going to require further adaption to the traditions and expectations of teaching in English schools. Given the many similarities between North American and UK contexts, this cautionary tale should perhaps give even more pause to those who propose importing approaches from much more exotic locales.
About the authors
Robert E. Slavin is a professor in the Institute for Effective Education at the University of York, Director of the Center for Research and Reform in Education at Johns Hopkins School of Education and the driving force behind the US-based Success for All Foundation, a restructuring program which helps schools to identify and implement strategies designed to meet the needs of all learners.
Mary Sheard is a research fellow in the Institute for Effective Education at the University of York. Her research interests are in literacy, teacher learning, pupil and student learning with technology, and learning with video representations.
Pam Hanley is a research fellow in the Institute for Effective Education at the University of York. Her interests include science education, continuing professional development, learning through small group discussion and the use of mixed methods in research.
Slavin RE (1995), Cooperative Learning: Theory, Research, and Practice (2nd Ed.). Boston: Allyn & Bacon.
Slavin R, and Lake C (2008), Effective Programs in Elementary Mathematics; A Best-evidence Synthesis. Review of Educational Research, 78 (3), 427–515.
Slavin RE, Lake C, and Groff C (2009), Effective Programs in Middle and High School Mathematics: A Best-evidence Synthesis. Review of Educational Research, 79 (2), 839–911.
Tracey L, Madden NA, and Slavin RE (2010), Effects of Co-operative Learning on the Mathematics Achievement of Years 4 and 5 Pupils in Britain: A Randomised Control Trial. Effective Education, 2 (1), 85–97.
This article is available to subscribers only. If you are an existing subscriber, please login. New subscribers may register below.