1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Was their classroom implementation observed to ascertain its level and quality? These problems appear even in some of the most sophisticated studies. Another problem is: In what kind of school environment was the study carried out? Is the school supportive of cooperative learning so that students appreciate that it is an inherent part of the school's norms, or is it viewed as an experiment not necessarily sanctioned by the school, and hence not to be taken seriously? Was cooperative learning adopted and practiced in one or two classes, did it extend to an entire cohort, was it adopted by the school as a matter of policy? Can educational research continue to disregard the organizational, school-wide context in which a given method is being practiced when it seeks to assess its effects? These problems can mislead investigators to reach over-generalized and even mistaken conclusions when the term cooperative learning was used without specification and the conclusions applied erroneously to cooperative learning in general. Educators who are not researchers can also be misled to think that the data support each and every method of cooperative learning, or, alternatively, fail to support all of the methods, when such is not the case. An error of this kind is not altogether surprising in behavioral research where, quite often, the investigators' main attention is focused on the dependent variables that they are measuring, while the independent variable that affected the students' or teachers' response is given short shrift in the research report. These and other critical problems are often illuminated by analytic reviews of research or by theoretical conceptions. Readers will be rewarded by paying close attention to the formulations of the conditions for productive learning groups offered by Elizabeth Cohen in her intensive review of research on cooperative learning (Cohen, 1994b). Noreen Webb's work on the quality of student interactions in small groups has inspired a substantial body of research (Webb, 1992). The implementation of cooperative learning in bilingual classrooms was explored recently in an entire issue of the journal Theory into Practice (Calderon & Slavin, 1999). One conception of the junior and senior high school as an organizational context in which cooperative learning can be practiced in an innovative and productive fashion is presented in a recent book (Sharan, Shachar, & Levine, 1999). The Quality of Student Interactions An observational study in 18 fourth-through-sixth grade classrooms from 4 schools in 2 districts (in California) was carried out to examine the relationship between patterns of student interactions in small groups and various academic and social outcomes (Battistich, Solomon, & Delucchi, 1993). The central finding of this study is that the effects of cooperative learning on students depends on the quality of their interaction. When group members were friendly, helpful to one another, concerned about each other's welfare, and worked collaboratively, there was a more positive classroom environment, the students liked school more, and displayed greater intrinsic motivation, and even performed better on standardized tests of achievement. In contrast, when group interactions were of a low quality, the effects of their group experiences were negative. The authors stress that none of the cooperative learning methods assume that simply having students work together in groups will have beneficial effects, and "many (methods-author) explicitly incorporate procedures for improving students' group interaction and management skills...." (page 27). This very topic was the goal of a study performed in Australia by Ashman and Gillies (1997). Students participating in this study were 192, 6th graders in 10 classrooms from 8 different schools. Groups receiving advanced training were "introduced to small group procedures through classroom discussion that encouraged students to identify skills relevant to the social