Appendix Table 1 provides the distribution of the county-level measures across the individual counties in the study, as well as the number of adolescent participants within them. On average, the seven urban counties with populations over 50,000 contain about twice as many youths in the study as those in the lower populus urban area and rural areas. Although several counties/court jurisdictions contained fewer than 10 youths in the sample, this is likely not a concern in the analysis. According to Maas and Hox (2004, 2005) and Snijders (2005), simulation studies suggest the primary concern regarding sample size in multilevel analysis is the between/group level because most multilevel studies contain very few groups compared to individuals within the groups. The present study contains a large group sample size (n = 73), which provides more confidence regarding the accuracy of the regression coefficients and standard errors estimated in the subsequent analysis. For the three low populus areas, there were twenty or more counties represented in the data. On average, the proportion of impoverished and Black youths was higher in the low populus rural and urban areas. The referral rates were higher in the urban areas, compared to the rural areas