TY - JOUR
AU - Stirratt, Mike
AB - Multivariate Behavioral Research, 36 (2), 155-183 Copyright © 2001, Lawrence Erlbaum Associates, Inc. Using Multilevel Modeling in the Evaluation of Community-Based Treatment Programs David Livert and David Rindskopf CUNY Graduate Center Leonard Saxe Brandeis University Mike Stirratt CUNY Graduate Center Health and social intervention programs targeted at neighborhoods, cities, and counties are becoming increasingly common (cf. Winick & Larson, 1997) and there is a concomitant need for systematic evaluation (Connell, Kubisch, Schorr & Weiss, 1995; O’Connor, 1995). Interventions across multiple sites require a different – multilevel – approach to testing hypotheses. Although multilevel models are increasingly being utilized (cf. Rindskopf & Saxe, 1998), actual application of such models to program assessment is complex and there are few examples. The goal of this article is to describe the application of multilevel statistical models to the evaluation of a multi-site community-based intervention and to explicate key issues in the design, analysis, and presentation of results. Since 1994, we have conducted the evaluation of Fighting Back (see Saxe, Reber, Hallfors, Kadushin, Jones, Rindskopf & Beveridge, 1997), a national demonstration program designed to test the feasibility of reducing substance abuse through coordination and expansion of community efforts (Jellinek & Hearn, 1991; Spickard, Dixon, 
TI - Using Multilevel Modeling in the Evaluation of Community-Based Treatment Programs
JF - Multivariate Behavioral Research
DO - 10.1207/S15327906MBR3602_02
DA - 2001-04-01
UR - https://www.deepdyve.com/lp/taylor-francis/using-multilevel-modeling-in-the-evaluation-of-community-based-6odzA9zipJ
SP - 155
EP - 183
VL - 36
IS - 2
DP - DeepDyve
ER -