School Attachment: Control for Economic Factors
Finally, we know that family income has been consistently associated with violence in children and is likely to be connected with school attachment. We found 7 studies reporting findings on school attachment and violence which controlled for family income or SES in some way. In 6 of the studies, negative associations between indicators of school attachment and violence remained robust in the face of a control for socioeconomics. Bellair et al. (2003) controlled for family income, Johnson (1979) included social class in his structural model, Loeber et al. (2005) controlled for low SES and “family on welfare” as well as “bad neighborhood” Ozbay and Ozcan (2006) controlled for “monetary strain” and income, Stewart and Simons (2006) controlled for family SES and neighborhood disadvantage, and Wright and Fitzpatrick (2006) controlled for family income and neighborhood median income. In this case, the only exception is the study by McNulty and Bellair (2003), who, like Bellair et al. (2003) above, used Add Health data. In our own analysis of Add Health data, we did find a significant, negative association between school attachment and violent behavior, controlling for family income.