Influences of friends and friendships on adjustment to junior high school

Merrill-Palmer Quarterly, Jan 1999 by Berndt, Thomas J, Hawkins, Jacquelyn A, Jiao, Ziyi

The ANOVA for general self-esteem revealed only a Cohort x Time interaction, F(2, 194) = 5.04, p

The ANOVA for teacher-rated behavioral problems had an N of 71 because, as noted earlier, ratings were unavailable for students in one elementary school. This analysis yielded a Sex x Cohort interaction, F (1, 67) = 6.27, p

Other ANOVAs suggested that the school transition had little effect on students' friendships. The ANOVA for friendship quality yielded no significant effects (ps > .05). No effects were significant when friendship stability was the dependent variable in an ANOVA with sex, cohort, and interval (Time 1-2 and Time 2-3) as factors (ps > .15). About half (49%) of students' best friendships remained stable across the transition interval (Time 1-2). Slightly more than half (55%) were stable across the posttransition interval (Time 2-3). The percentages for the two intervals do not differ significantly (p > .20).

Continuities in Adjustment and Friendship After the Transition

If the school transition affected some students more than others, it could alter the pattern of individual differences in adjustment and friendship quality. To evaluate this possibility, correlations for the continuity of individual differences were computed. As Table 3 shows, all correlations were above .45 and several were above .70 for the transition and posttransition intervals. Similar correlations were reported in previous studies of the junior high transition (e.g., Eccles et al., 1989; Simmons & Blyth, 1987).

More important than the magnitude of the correlations are their relative values for the two intervals. A test described by Steiger (1980) showed that the correlation for sensitivity-isolation was significantly higher (p

Changes in grading systems and in teachers between elementary and junior high school may partly explain the differences in the correlations for grades and behavioral problems. However, no changes in measurement can explain why self-reports of social competence and peer nominations for sociability-leadership showed lower continuity across the transition than during seventh grade. Apparently, substantial changes in peer relationships occurred when students moved to junior high, but these relationships stabilized during the first semester in the new school.

The Influences of Friends and Friendships: Analysis Plan

Hierarchical regression analyses were used to examine the degree to which the changes over time in students' adjustment were affected by the quality and stability of their friendships, by their friends' adjustment, and by the interactions of these variables. Two sets of regression analyses were done. The first set tested the hypothesis that students' adjustment improves over time if they have high-quality friendships, but perhaps only if these friendships are stable. More generally, these analyses examined the effects of friendship quality, friendship stability, and their interaction. The analyses included all students in the sample, and the scores for friendship quality and stability were based on all the friendships that students described.

The second set of analyses tested the hypothesis that students' adjustment improves over time if their friends are high in adjustment. The same analyses tested the more complex hypothesis that the influence of the friends' adjustment is greater when these friendships are higher in quality and stability. More generally, these analyses examined the effects of friends' adjustment both by itself and in interaction with friendship quality and stability. The analyses included the subsample of students who could be matched with one or more friends. The scores for friendship quality and stability used in these analyses were those for matched friends only.