The educational potential of peer collaboration in middle childhood was examined in this study in two ways. First, the pretest-posttest change of same sex dyads was compared. These pairs worked together for six sessions, in each of two years, on several tasks. Secondly, the social interactions of the dyads videotaped during the peer sessions were analyzed. Data were first collected on fourth graders, and again a year later when they were in fifth grade. In Year 1 (1985), two experimental groups and one control group were arranged. Experimental Group 1 were 40 fourth graders who worked in pairs to solve math problems. Experimental Group 2 included 40 children working on spatial problems. There were six weekly training sessions, during which the children were presented with increasingly more difficult problems in either the math or spatial area. The control group was 42 children who received only the same pretest and posttests as the two experimental groups. In Year 2 (1986), all of the Year 1 children were given a problem in physical equilibrium (Siegler's "balance beam") to work on. They worked in dyads over a six week period. A new control group of 42 fifth graders who were pretested and posttested on the balance beam problem but who did not participate in the learning sessions was added. Materials used during the peer math sessions were missing addend and multiplication number problems, word problems, ratio problems, and proportion problems. Materials for the spatial reasoning sessions were perspective-taking problems, construction problems, and map-drawing problems. The six balance scale sessions each consisted of 20 incrementally more difficult problems. The Murray Center holds the following raw data from this study: pretests/posttests, materials generated by children during the sessions, videotapes of sessions 1, 4, and 6 in Years 1 and 2, and transcripts of sessions 1 and 6 in both years. Test scores, coded transcripts, summary interaction codes, and sequential codes are available in computer-accessible form.
