| This study investigated the effect of gender and group gender composition on interaction patterns and attitudes toward their cooperative learning experience of seventh- and eighth-grade students working with the multimedia program Loess Hills Interactive. One hundred twenty-seven middle school science students (60 males, 67 females) from two schools in Iowa participated in the month-long study. Students were randomly assigned to cooperative learning groups of 3, 4, and 5 with the following gender compositions: (1) same-gender, (2) mostly-male, (3) mostly-female, and (4) equal-gender. Trained raters coded and tallied interaction frequencies for each student from videotapes, using a Peer Interaction Scale developed from pilot study data. The interaction categories included: (a) path/pace, (b) task, (c) socio-emotional, (d) technical, (e) off-task, and (f) uncodable. An attitudinal survey, administered at the end of the four weeks, measured five factors: (1) positive emotional reaction to the group, (2) presence of helping behaviors in the group, (3) preference for working alone, (4) lack of helping behaviors in the group, and (5) preference for small-group learning.; Two-way analyses of variance were performed on the verbal interaction and attitudinal data to determine if significant differences occurred between males and females in groups of varying gender composition. Student gender did not have a significant effect on either interactions or attitude. However, when group and school were the independent variables in two-way analyses of variance, significant effects were found for group composition on three peer interaction categories (total, path/pace, and task) and for school on five categories (total, path/pace, task, socio-emotional, and off-task). Students in mostly-female groups scored significantly higher than mostly-male or equal-gender groups in their preference for small-group learning. Pearson product-moment correlations determined that no significant correlations occurred between the interaction and attitudinal data. Differences in students, teacher involvement, and implementation between the two schools may have influenced interaction and attitudinal data. |
