One of the most frequently cited reasons for conducting a meta-analysis is the increase in statistical power that it affords a reviewer. This article demonstrates that fixed-effects meta-analysis increases statistical power by reducing the standard error of the weighted average effect size (T.) and, in so doing, shrinks the confidence interval around T.. Small confidence intervals make it more likely for reviewers to detect nonzero population effects, thereby increasing statistical power. Smaller confidence intervals also represent increased precision of the estimated population effect size. Computational examples are provided for 3 effect-size indices: d (standardized mean difference), Pearson's r, and odds ratios. Randomeffects meta-analyses also may show increased statistical power and a smaller standard error of the weighted average effect size. However, the authors demonstrate that increasing the number of studies in a random-effects meta-analysis does not always increase statistical power.

This study investigated age changes in risk perception and unrealistic optimism. Teenagers (« = 376) and parents (n = 160) evaluated the risk of experimental, occasional, and regular involvement in 14 health-related activities (e.g., getting drunk). Respondents also evaluated their comparative chances of encountering the leading causes of morbidity and mortality. Compared with adults, teenagers minimized the perceived risk of experimental and occasional involvement in health-threatening activities. Notably, teenagers were less optimistic about avoiding injury and illness than were their parents, and teenagers at greatest risk for such misfortunes were the least optimistic about avoiding them. These findings do not support traditional explanations of adolescent risk taking. The implications of these findings for understanding and preventing health-damaging behavior among adolescents are discussed.

Using the Sentence Completion Test for ego development, we studied several cohorts of students between 1971 and 1979 at a technological institute (Tech) and between 1974 and 1979 at a predominantly liberal arts university (MU). Ego level tended to rise slightly except among women at MU, for whom there was a slight but consistent loss. This particular finding challenges one assumption of a widely accepted version of Piagetian theory: that stage development is irreversible. Women tended to enter MU slightly ahead of men in ego level, but left at the same level. Contrary to expectation, men and women appeared to gain more at Tech than at MU; the difference was significant only for women.

Size and stability of sex differences in personality growth throughout adolescence and adulthood are examined. Studies using the Washington University Sentence Completion Test of ego development served as the primary source data. Sixty-five studies (encompassing more than 9,000 Ss) generated 113 independent effect sizes. Sex differences in ego development were moderately large among junior and senior high-school students (female advantage), declined significantly among college-age adults, and disappeared entirely among older men and women. Sex differences were relatively stable during early and middle adolescence. The greater maturity displayed by adolescent girls is not an artifact of superior verbal abilities. Sex differences in ego development were more than twice the magnitude of differences in vocabulary skills (Hyde & Linn, 1988). The present findings, together with findings from reviews of sex differences in moral judgment, aggression, and empathy, suggest that adolescent girls achieve developmental milestones earlier than boys, a difference that declines with age.

Previous research has examined body dissatisfaction and pressures toward thinness among collegeage and adult women, demonstrating greater dissatisfaction among women than men. Little is known about when such sex differences arise. The present study replicated the procedure used by Fallen and Rozin (1985) to assess body-size preferences in a sample of 288 female and 283 male adolescents aged 10.5 to 15 years. Both sexes revealed a small degree of body-figure dissatisfaction relative to their chosen ideal, but neither sex rated their own figure as significantly different from the size considered most attractive to the opposite sex. Both male and female adolescents held distorted perceptions of opposite-sex preferences. Girls showed a bias toward thinness; boys revealed a bias toward larger figures. The latter bias was associated with pubertal development. Fallen and Rozin (1985) found that college women perceive their figures to be heavier than the figure they consider most attractive to men, and they view as ideal a figure that is thinner than either of the latter assessments. In contrast, college men display no differences in the perceptions of their own figure, ideal figure, and figure they consider most attractive to women. These and other findings (e.g., Dwyer, Feldman, & Mayer, 1970;Klesges, 1983) indicate a greater pressure for thinness among women than among men. This pressure has been implicated as a contributing factor to the higher incidence of eating disorders in women (Garner, Garfinkel, Schwartz, & Thompson, 1980;Striegel-Moore, Silberstein, & Rodin, 1986). It is not known how or at what point youngsters incorporate such pressures.This study examines the developmental roots of body-figure preferences by replicating Fallon and Rozin's (1985) procedure with a sample of early adolescents. This age group is of particular interest because of the dramatic physical changes that accompany puberty and their potential impact on the establishment of a body ideal. MethodSubjects were 283 boys and 288 girls recruited from an urban middle school (160 sixth graders, 176 seventh graders, and 235 eighth graders). The mean age for both boys and girls was 13.0 years (SD = 0.96).