“…Additionally, it is expected that the influence of legal factors across county may vary by race of the juvenile. Accordingly, these data were partitioned by county and subsequently by race for each county in order to produce separate estimates for the urban and suburban county, permitting a comparison of the coefficients for each group (Albonetti, 1987; Myers, 1985). The equality of the coefficients across models was tested using the z test (Paternoster, Brame, Mazerolle, & Piquero, 1998).…”