Study 1: Peers

Method

Subjects. Same- and mixed-gender randomly assigned peer dyads were observed (N=36 pairs). The mean age of the children was 3.7 years (range, 2.4 to 4.9). The children were pided at the median age (3.5 years) into older and younger groups. The younger group's average age was. 2.0 years (range, 2.4 to 3.5). The older group's average age was 4.5 years (range, 3.6 to 4.9). Equal numbers of girls and boys participated; eight girls and eight boys comprised the younger group, ten girls and ten boys comprised the older group (see Table 1). Twelve of the children were singletons fifteen had older siblings, and eleven had younger siblings. All of the children carne from middle- to upper-middle-class intact families; the ethnic majority was Caucasian.

Procedure. Each dyad was observed with the experimenter unobtrusively present in a laboratory playroom during 2 ten-minute sessions. For each ten-minute session 4 toys were present, 2 judged to be masculine and 2 judged to be feminine. In one session all toys were typically inpidualistic and in the other session all toys were typically social in function. Order of toys was counterbalanced across dyads (because order was not a factor, it is not included as an independent variable). The children were allowed to play together as naturally as possible. The sessions were videotaped and all conflicts involving toys were identified.

The behavioral code used to analyze the videotapes for conflicts, precipitating events and protests is provided in Table 2. Coder reliability was checked on 33% of the conflicts. Two coders agreed that conflict was occurring on 77% of occasions. The kappa coefficient was .66 overall for precipitations and .61 overall for protests.

Results

Description of analyses. To examine the effects of the independent variables on features of conflict 2 (same-or mixed-gender dyads) by 2 (age) by 2 (toy function) mixed-model analyses of variance (ANOVA) were computed, with toy function as the repeated measures factor. One dyad had no toy-related conflicts and was therefore dropped from the analyses. For type of precipitating and protest events, multiple dependent variables were analyzed. Because of small cell sizes, MANOVAs could not be conducted. Therefore, for each a series of ANOVAs was conducted, using a Bonferonni adjustment to set the alpha level in each case, with p=.02 for protests arid p=.01 for precipitating events (Hertzog & Royine, 1985) Where appropriate, chi-square analyses were, computed to examine distributions across categories. Whenever percentages were analyzed, arcsine transformations were used (Alder & Roessler, 1972).

Frequency and duration. In total, 114 conflicts were identified. On average each dyad had 4.11 conflicts (range 0-10). Each dyad spent an average of 39.47 seconds in conflict (range 0-282 seconds), approximately 3% of total contact time per dyad. Forty-six percent of conflicts occurred in mixed-gender pairs. and 54% in same-gender pairs (See Table 3). There were no main effects or interactions for frequency and duration analyses.

Nature of protests. The proportion of communicative acts (30%), direct acts (34%), or a combination of the two (36%), did not differ from a chance distribution (see Table 4). For communicative acts, a main effect of age, F(1,31) = 12.21, p=.001, indicated that older peers protested verbally (M=.94) more often than younger peers (M=.43). There was a nonsignificant trend (p=.04) for younger children to protest directly (M=1.06) more often than older peers (M=.43).

Nature of precipitating events. Precipitating events were most often direct (69%), followed by communicative (27%) and combination acts (4%), chi2(2, N=144)=96.54, p=.001(see Table 5). Again, for communicative acts, a main effect of age, F(1,31)=5.50, p=.02, indicated that older peers were more likely to precipitate conflict verbally (M=.84) than were younger peers (M=.22). For direct acts, a main effect of age, F(1,31)=5.80, p=-.02, indicated that younger peers were more likely to precipitate conflict physically (M=1.88) than were (older peers (M=1.04). There was a trend, F(1,31)=5.05, p=.03, for same-sex pairs to be more physical (M=1.84) m their precipitating behavior than mixed-sex pairs (1.07).