kitchen table math, the sequel: Evidence for cooperative learning?

Monday, August 10, 2009

Evidence for cooperative learning?

In defense of requiring his students to work in groups, a teacher (posting at education.change.org) recently cited the following, from Johnson, Johnson, and Stanne (2000):
Cooperative learning has been around a long time (Johnson, 1970; Johnson & Johnson, 1989, 1999). It will probably never go away due to its rich history of theory, research, and actual use in the classroom. Markedly different theoretical perspectives (social interdependence, cognitive-developmental, and behavioral learning) provide a clear rationale as to why cooperative efforts are essential for maximizing learning and ensuring healthy cognitive and social development as well as many other important instructional outcomes. Hundreds of research studies demonstrate that cooperative efforts result in higher individual achievement than do competitive or individualistic efforts. Educators use cooperative learning throughout North America, Europe, and many other parts of the world. This combination of theory, research, and practice makes cooperative learning one of the most distinguished of all instructional practices.
Here, cooperative learning is defined as follows:
Cooperative learning exists when students work together to accomplish shared learning goals (Johnson & Johnson, 1999). Each student can then achieve his or her learning goal if and only if the other group members achieve theirs (Deutsch, 1962).
You can read the whole meta-analysis here.

Does anyone have any thoughts on how reliable this meta-analysis is (it seems to be unpublished, only appearing on the University of Minnesota Cooperative Learning Center website, and finds Johnson & Johnson's own method, Learning Together, most effective), or whether there's any research on cooperative learning that contradicts its conclusions?

18 comments:

le radical galoisien said...
This comment has been removed by the author.
le radical galoisien said...

My Singapore math teacher used groups occasionally (frequently?). But she always knew what she was doing, and she always controlled the lesson.

And group work was almost always done in the context of competition. You had to do better than the other groups if you wanted to win the prize (ice cream sandwiches for recess, or something); think PSLE math with required working, except on a transparency and on a 3-minute time limit.

One thing that I think is a massive waste of time in the Singapore curriculum is group project work. I don't know if they're trying to make us engineers at ten-years-old, but the teachers almost never really controlled the sessions. I always felt "the thing we're designing is so pointless! A new kind of potato chip can?? Really??"

And for 10-year-olds, the teachers never offered any guidance on consensus-decision-making. It also sucks if you have a creative idea, and you try to address everyone's ideas and propose some sort of synthesis, because then your group thinks you're being bossy (especially if you come with an American accent) and chaos erupts and the teacher thinks it's all your fault. =(

Catherine Johnson said...

I am going to consult Hattie right this minute.

Catherine Johnson said...

OK, Hattie is very interesting.

Hattie says that both cooperative and competitive learning are more effective than "individualized" instruction, which I assume to be akin to differentiated instruction.

Catherine Johnson said...

Teaching LD puts Cooperative Learning under "Use Caution" (i.e.: "practices for which the research evidence is incomplete, mixed, or negative).

They are talking about kids with learning disabilities specifically, not general education students.

le radical galoisien said...

Couldn't you combine both competitive and cooperative aspects?

Catherine Johnson said...

The Hattie book is going to be quite useful.

Basically, what his metanalysis shows, is that peers are important, PERIOD.

I believe that; Albert Bandura showed quite some time ago that classic trial-and-error learning isn't the major mode of learning for any creature (I believe). The requirements of survival mean that the baby antelope has to learn to stay away from lions by watching other antelope stay away from lions.

And, of course, Irene Pepperberg showed that while it's impossible to train birds using classic reinforcement training, when you introduce a 'third party' into the learning situation (called the 'model/riveal,' parrots turn out to be able students.

Here's an interesting page on Alex & model/rival training.

Catherine Johnson said...

From Teaching LD (pdf file):

Whereas CL strategies typically involve two or more students working together to accomplish an assigned task, it is not synonymous with “group work.” Johnson and Johnson (1994) identified five elements critical to maintaining structure and student involvement in CL: (1) positive interdependence, which means students realize that group performance depends on the contributions of each member; (2) face-to-face promotive interaction, wherein students encourage and facilitate each other’s efforts to achieve; (3) individual accountability; (4) the use of interpersonal skills; and (5) group processing, which refers to groups’ reflections on how well they are functioning. Researchers emphasize that teaching students such interpersonal behaviors and monitoring their use are critical to the success of CL (e.g., Goor & Schwenn, 1993; Johnson & Johnson, 1992).

Catherine Johnson said...

Another passage from Hattie:

"All of the many meta-analyses by the Johnsons and their colleagues show high effect sizes, whereas the others hover around the small to medium effects.

[snip]

"Johnson and Johnson (1987) argued also that cooperation was most effective among adults...

[snip]

"Cooperative learning is more effective in reading (Hall, 1988, d=0.44) than in mathematics (d=0.01), and Johnson et al. (1981) found that for rote decoding and correcting tasks, cooperation does not seem to be superior."

People should correct me if I'm wrong, but my sense of an effect size of .01 in an education study is that we should assume it is functionally no different from 0.

Anonymous said...

Peers are important because Peers ARE the culture.

When you say
"it's the culture, stupid", you often mean the school culture. But really, those cultures--school culture, etc., are just incubators or the growing medium for the actual culture in which people learn.

Katharine Beals said...

Thanks, Catherine, for your digestion of what Hattie has to say on this. I'll have to read the book! (Just finished another book I learned about here: Education Myths)

Anonymous said...

What's missing from the meta-analysis from U Minn is context. There's no sense of what types of learning tasks are being used to compare cooperative vs. individual learning. My experience as a parent and teacher is that cooperative learning is s technique that can be useful for specific types of learning, but that it is inappropriate for the majority of learning objectives.

Catherine Johnson said...

My experience as a parent and teacher is that cooperative learning is s technique that can be useful for specific types of learning, but that it is inappropriate for the majority of learning objectives.

Anonymous - if you're around & have the time - could you fill us in a little more?

Catherine Johnson said...

I had a fascinating conversation with one of my most perceptive friends yesterday (who I don't see often, unfortunately).

She instantly sparked to the 'cooperative learning' question in terms of ... the peer/competitor effect. In cross-country coaching, she says (which our district does extremely well), you put slower runners with faster runners. People always run faster when they're running with a partner, and a major mode of training is to put slower runners with faster runners.

This is planned out carefully; it isn't a matter of willy-nilly tossing the slow kids in with the fast kids. The coach picks which 3 slower runners, say, will run with the fastest kids.

I asked twice whether the slower kids slow down the faster kids & she said 'no.' (Which I believe because, as I said, track coaching is something this district does very well.)

The faster runners speed up the slower runners.

She said it absolutely makes sense to her that when you're doing 'cooperative' learning correctly you're going to have the 'pacing' effect that coaches use.

I think I'll repeat that when Hattie says 'group' effects are stronger than 'individualized instruction,' he means cooperative learning AND competitive learning: i.e. competition between or amongst groups or teams.

Now that I've thought of it this way, it makes a lot of sense. C's classes at Hogwarts have 28 kids & are taught via whole-group instruction, and they are jolly. All last year I had the sense that there is something incredibly 'friendly' about a large, cohesive class -- and I guess what I was perceiving was a group effect.

People are social animals - maybe teens especially!

At Hogwarts you have all boys, all dressed in school uniforms, and all attending fairly large classrooms listening to the same thing at the same time ---

It reminds me of the time, years ago, when I got a last minute ticket to a Bruce Springsteen concert. It was an open-air stadium & my friend and I were just a couple of rows down from the top --- and at one point it hit me: there is a particular kind of pleasure to be had when 30,000 people are thinking and doing the exact same thing at the exact same time and in the exact same place.

Catherine Johnson said...

btw, I also found an explanation of 'peer tutoring' - which turns out to be NOTHING like what I had thought.

It's a specific concept, invented by (I think) a specific school, and it's a form of competition.

I'll get that posted.

Anonymous said...

Catherine, the effect you note with training for cross-country (which I've also observed) is quite different from "cooperative learning." In that sports situation, children/teens are not being expected to collaborate on a complex, multi-step, somewhat undefined project. All they're expected to do is run, with the example of the faster runners having a motivating effect on the slower ones.

"Cooperative learning," as the term is used in instructional settings (and vaguely alluded to in the meta-analysis referenced above) usually means a team research and presentation process where each student performs a part of the whole. This can range from each student actually doing some sort of research (from interviewing someone to looking something up on the internet to finding pictures of something, etc etc) to a more differentiated process where one student does research, one facilitates a conversation where research findings are discussed, one records, one writes the results down, and one presens verbally (for example). As you can see, this can be a worthwhile experience ONCE IN A WHILE. It's not an effecient way to learn content or even learn research skills. And the devil is in the details, especially when roles such as "provider of refreshments" or "timekeeper" are added in -- those students are not learning anything remotely related to the task at hand.

The fact that a learning module involves whole-group instruction is actually the reverse of cooperative learning, to most of its advocates. There, the relevant unit is a group, but the students are not cooperating with each other, they're consuming the teacher-provided instruction. As another "guerilla instructivist," I think that's a good thing, as long as the instruction gets the job done.

Does that respond to your questions?

le radical galoisien said...
This comment has been removed by the author.
le radical galoisien said...

It seems to me that the best way to use the method would be to use competing groups. Thus, the researcher is competing against other researchers; the presenter is competing against other presenters.

And of course, the competitive aspect will have to be emphasised for the competitive aspect to have full effect -- that is, competing researchers should have ample chances to see what other researchers are doing.

If performance is reviewed (between groups) every 5 minutes, rather than at the end of the week, that would motivate groups to outcompete each other, because they can see they are falling behind. Just like slow runners are basically getting performance reviews every time they take a breath.