Risk and resilience in the urban neighborhood: Predictors of academic performance among low-income elementary school children

Merrill-Palmer Quarterly, Apr 1999 by Shumow, Lee, Vandell, Deborah Lowe, Posner, Jill

In order to examine potential compensatory effects associated with intra-individual, familial, and community factors, a series of hierarchical multiple regressions were conducted. In the first equation, the block consisting of family structure, race, maternal education, and child gender was used to predict child school performance. The blocks containing the individual, family, and community factors were then tested in three separate equations. The first equation considered whether the addition of intra-individual factors provided a significant increment to R2. As shown in Table 5, the block of intra-individual factors contributed an additional 19% to the model. Inspection of the betas indicates that impulse control (p

The next regression equations considered the addition of the familial resilience block to the demographic controls. The block of familial factors contributed an additional 21% to the model predicting academic performance. The betas indicated that parental school involvement was uniquely associated with better school adjustment, whereas family emotional support was not.

In the third regression the community resiliency block was added to the individual demographic controls. The community variables contributed 3% beyond the control factors, a nonsignificant increment.

Finally, we conducted a simultaneous multiple regression in which the demographic, intra-individual, and familial factors that were significant in the prior models were entered into a full model. This exploratory equation, displayed in the last column of Table 5, contained child gender, race, maternal education, social problem-solving skill, impulse control, academic self-esteem, and maternal school involvement. As can be seen, impulse control and maternal school involvement were unique predictors among the resiliency factors considered in this equation.

DISCUSSION

Children's academic performance in fifth grade was found to be negatively associated with an indicator of neighborhood risk even after controlling for demographic markers of familial risk. This relation was not evident in third grade when controlling for individual family characteristics. These findings are consistent with the results of Entwisle et al. (1994) relating neighborhood socioeconomic factors with changes in children's mathematical reasoning between third and eighth grades. Our findings are also consistent with research of neighborhood influences on academic problems among one sample of children attending a middle school (grades 6-8), but not among a different sample of 9-year-old elementary school children (Halpern-Felsher et al., 1997). Prior researchers have associated neighborhood risk with adolescent school outcomes (Brooks-Gunn et al., 1991; Connell et al., 1994; Crane, 1991), but not with first graders' academic skills after controlling for family factors (Chase-Lansdale & Gordon, 1996). Collectively, these studies suggest that neighborhood influences may become stronger at the end of middle childhood.