• By Dartington SRU
• Posted on Tuesday 05^th August, 2014

How the culture of the classroom matters

Cooperative learning has been repeatedly shown to be an effective way to improve academic achievement – in the US, at least. When cooperative learning strategies for mathematics were tested in a randomised controlled trial in the UK, the results were, as the researchers put it, “unexpected and disappointing.”

Why might this be the case?

There is debate among scholars as to why a programme shows positive results during one test, but then fails during another.

Some say it has to do with whether or not the developer was involved: evaluations that include the programme developer tend to show better results than those that don’t. But in this case, the developer was involved.

Others argue that developers try to replicate a programme prematurely or try to take it to scale before it is really ready. In the case of the cooperative learning programme, this was not the first replication. Nor was it carried out on a much larger scale than previous studies.

The answer here is likely much simpler and intuitive. That is, the replication of a programme ought to actually replicate it. But this “replication” suffered from two problems.

First, the cooperative learning programme had been altered to fit the UK context. This contextual adaptation, however, changed a core component of the programme and likely contributed to its failure to produce significant results.

Second, and perhaps even more important, the culture of UK classrooms – oriented toward individualised learning – makes it difficult for teachers to feel comfortable with cooperative learning strategies. Indeed, a quarter of the schools assigned to the intervention dropped out. And of the teachers who remained, less than half implemented the programme faithfully, researchers said.

Cooperative learning structures classes around small groups that work together so that each individual’s success is dependent on the group's success. Students are assigned to heterogeneous learning teams where the students are of mixed abilities, gender, and ethnicity.

The teacher presents a lesson, and then students work within their teams to make sure that all team members have mastered the lesson. All students then take individual quizzes on the material, at which time they may not help one another.

The teams are rewarded for the quiz scores of the group as a whole. Students’ quiz scores are compared to their own past averages, and points are awarded on the basis of the degree to which students meet or exceed their own earlier performance. These points are then summed to form team scores. The teams are then recognised as whole for their performance, either with a certificate or other types of rewards.

The cooperative learning model used in the UK test was developed by psychologist Robert Slavin and colleagues at the Baltimore-based Success for All Foundation. The programme is referred to as Power Teaching Mathematics in the US and Power Maths in the UK.

During the 2008-09 school year, researchers experimentally tested Power Maths with children ages 8-10 in the Year 4 and 5 classrooms of 34 UK-based elementary schools. In contrast to studies conducted elsewhere, students’ mathematics achievement did not improve on average as a result of the introduction of cooperative learning.

The students who received Power Maths and the students who received traditional mathematics teaching had statistically equivalent math achievement scores at the start and at the end of the study, indicating that the programme had no effect.

A major difference between how the programme was applied in this test and how it was applied elsewhere was the decision not to use group scoring.

The researchers noted that UK teachers were under pressure to use “differentiated instruction” – which means teaching students at the level in which they are ready to learn (that is, one lesson plan for more advanced students and another for less advanced students).

As such, many teachers were inclined to group students of the same ability together and “were very uncomfortable teaching mixed-ability groups,” says the research team. Because of this discomfort, group scoring was replaced with randomly selecting individual team members to represent their teams.

The problem is that group scoring, which provides group goals and individual accountability, is considered essential to the outcomes of cooperative learning. By abandoning group scoring, the programme altered one of its core components and, perhaps not surprisingly, altered its effectiveness.

And the UK classroom culture of individualised learning may also have been behind the high dropout rates, which also contributed markedly to the programme’s lack of success, the researchers say. Of the 20 schools originally assigned to the Power Maths programme, only 15 stuck with the programme for a full year.

To make matters worse, of the 24 teachers who remained in the programme, only 10 were rated as having delivered Power Maths faithfully.

Sharing effective programmes across national boundaries remains a crucial way to make evidence-based interventions more widely available. In making these border crossings, programme developers face a paradox. As this study shows, they must take account of the culture and preferences of the new organisations that will put the programme into practice, or risk high dropout rates. At the same time, core components must remain faithful, or the programme may not work as intended.

Tracey, L., Madden, N.A., & Slavin, R.E. (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