CalStateStudyOfGroupLearning

Part 2 in a mini-series on a review of quality math ed research articles.

Part 1: CalStateStudyIntro

The most surprising thing, to my mind, in the Cal State Study is its strong endorsement of cooperative group learning. The review included eleven studies of group learning, all with positive findings for the use of structured group learning.

The Cal State Study defines "conventional mathematics instruction" as being characterized by teacher explanation of the new material, followed by independent workbook activity. The study makes the strong claim that the conventional approach has absolutely no theoretical support, and is discredited by the totality of the studies in the review that examine cooperative group learning methods.

Actually, this is the one place I've read in the report so far where I felt I might be encountering just a whiff of reviewer bias, particularly in the glowing interpretation of these findings.

But the evidence might really be there. Carefully constructed group learning opportunities beat conventional teaching methods. Go figure.

The phrase 'carefully constructed' is a point to dwell on. Simply sticking kids together in groups to do their homework conferred no benefit. The group learning environment had to be structured, i.e. the nature of the kids' interactions had to be controlled by external reinforcement systems. In particular, higher performing kids had to be motivated to help lower performers.

To some degree, it didn't even matter what the system was, but it had to be present. One study compared cooperative vs. competitive group reinforcement systems: an example of a cooperative system is one in which each kid in a group gets the average of their individual grades, whereas grouping the kids into competitive teams is a competitive reinforcement system. Both reinforcement systems were shown to confer equal benefits above the conventional teaching style.

The reviewers note that "a particularly interesting aspect of these studies on cooperative work is that all of them included students who were at risk in some sense, either by virtue of being inner city urban students, and/or low SES students, and/or students identified as having special learning needs."

While the reviewers seem rather pleased with this study design, to my mind it weakens the broad applicability of the results. I rather wish they'd done these studies with the most typical bunch of learners they could find, since the presence of special needs can really skew a kid's learning style. Kids who are starved for attention, interaction, or positive reinforcement for learning, might benefit much more from learning environments that include these things than a typical learner would.

Also see: CalBoardOfEdStudyPart2, EdResearch


California study intro
California state study of group learning
California Board of Ed study part 2
education research - peer reviewed studies - chart

Back to main page.