This paper describes an improved method for performing statistical comparisons among experimental groups. This technique, termed multi-response permutation procedures (MRPP), is similar in purpose to the t test and one-way analysis of variance F test. However, in contrast to these, the new method features very relaxed requirements on the data structure, is easily applied to multivariate problems, and makes it possible to relate the analysis visually to the perceived data space. The MRPP test statistic is based on the within-group average of pairwise distance measures between object response values in a euclidian data space. The null distribution of the test statistic is based on the collection of all possible permutations of the objects into groups having specified sizes. For large group sizes, this distribution is approximated by a continuous distribution satisfYing three exact moments.The advantages and applications of MRPP are illustrated using both artificial examples and empirical data on total August standing crop in mixed prairie following an October wildfire. The MRPP analyses of the empirical data revealed that there were no differences in standing crop between burned and unburned areas after the first postfire growing season, but that after two growing seasons, standing crop was significantly greater in the previously burned areas.

We used data from the Project on Human Development in Chicago Neighborhoods to examine the extent to which individual, family, and contextual factors account for the differential exposure to violence associated with race/ethnicity among youths. Logistic hierarchical item response models on 2344 individuals nested within 80 neighborhoods revealed that the odds of being exposed to violence were 74% and 112% higher for Hispanics and Blacks, respectively, than for Whites. Appreciable portions of the Hispanic-White gap (33%) and the Black-White gap (53%) were accounted for by family background factors, individual differences, and neighborhood factors. The findings imply that programs aimed at addressing the risk factors for exposure to violence and alleviating the effects of exposure to violence may decrease racial/ethnic disparities in exposure to violence and its consequences.

Although researchers consistently demonstrate that females engage in less criminal behavior than males across the life course, research on the variability of the gender gap across contexts is sparse. To address this issue, we examine the gender gap in self-reported violent crime among adolescents across neighborhoods. Multilevel models using data from the Project of Human Development in Chicago Neighborhoods (PHDCN) indicate that the gender gap in violent crime decreases as levels of neighborhood disadvantage increase. Further, the narrowing of the gender gap is explained by gender differences in peer influence on violent offending. Neighborhood disadvantage increases exposure to peer violence for both sexes, but peer violence has a stronger impact on violent offending for females than for males, producing the reduction in the gender gap at higher levels of disadvantage. We also find that the gender difference in the relationship between peer violence and offending is explained, in part, by (1) the tendency for females to have more intimate friendships than males, and (2) the moderating effect of peer intimacy on the relationship between peer violence and self-reported violent behavior.

No abstract