Multivariate Analyses of School Attachment

One might suspect that confounding factors play a role in this unexpected association. We might make the case that factors such as academic achievement (which, as we have seen, is consistently associated with violence and which we would expect to be strongly associated with school attachment), parent education, and family SES would be important control variables. First, we will discuss the findings in the large multivariate models. Then, we will focus on studies employing controls for those specific constructs.

If we focus on the multivariate analyses overall, we find very strong and consistent support for an association between violence and school attachment in the predicted direction. The association has been identified in studies controlling for delinquent peers (e.g., Bellair et al., 2003; Friedman & Rosenbaum, 1988; Johnson, 1979; Ozbay & Ozcan, 2006; Sprott et al., 2005; Stewart & Simons, 2006), attachment to parents or parent-child relationship (e.g., Bellair et al., 2003; Friedman & Rosenbaum, 1988; Gardner & Shoemaker, 1989; Johnson, 1979; Ozbay & Ozcan, 2006; Wright & Fitzpatrick, 2006), parental monitoring (e.g., Benda, 2005; Ozbay & Ozcan, 2006; Wright & Fitzpatrick, 2006), various forms of participation in sports, clubs, religious organizations (e.g., Friedman & Rosenbaum, 1988; Gardner & Shoemaker, 1989; Ozbay & Ozcan, 2006; Wright & Fitzpatrick, 2006), neighborhood disadvantage or neighborhood problems (Bellair et al., 2003; Gottfredson, McNeil, & Gottfredson, 1991; Stewart & Simons, 2006), alcohol or drug use (Bellair et al., 2003), abuse victimization (Benda, 2005; Stewart & Simons, 2006), miscellaneous psychological factors (Benda, 2005; Reis, Trockel, & Mulhall, 2007), and cumulative risk (Sprott et al., 2005).

 
Source
< Prev   CONTENTS   Source   Next >