Persistently Safe Schools: The National Conference of the HAMILTON FISH INSTITUTE ON SCHOOL AND COMMUNITY VIOLENCE MEASURING CHANGE IN COLLABORATION AMONG SCHOOL SAFETY PARTNERS Bruce B. Frey Jill H. Lohmeier Steve W. Lee Nona Tollefson Mary Lea Johanning University of Kansas Abstract The federal Safe Schools, Healthy Students grants awarded to hundred of cities and school districts over the last four years include evaluation mandates for a variety of outcomes. Typically, sustainability of any initiative is an explicit or implicit objective; it is hoped that collaborations formed for the purposes of implementing the grants will remain after the funding agencies have ended direct involvement and oversight. From the perspective of evaluators, however, assessing collaboration is often difficult. Models of collaboration among agencies, groups, and community stakeholders are notoriously difficult to translate into valid and reliable instruments that can be used to measure meaningful change in the level and pattern of collaboration. This paper describes a model of collaboration, presents an instrument for its assessment, and suggests a method of graphical display that captures the sometimes elusive nature of collaboration. In designing the evaluation of a local Safe Schools, Healthy Students initiative, we reviewed the literature on collaboration among public and private agencies and chose a model which identifies fairly well-defined levels or stages of collaboration (Borden and Perkins, 1998) and developed a survey instrument which would measure grant partners perceptions of their collaboration with each other. A method of visual display which shows collaborative links and their strength among partners was adapted from a method presented by Cross (2003). Early evidence of scale reliability for the Levels of Collaboration Scale is presented and issues of validity are discussed. Introduction Safe Schools, Healthy Students grants, first funded in 1999 through the federal Departments of Education (USED), Justice (DOJ), and Health and Human Services (HHS), have been awarded to hundreds of cities and school districts over the last five years and include evaluation mandates. An increased level of collaboration is frequently an explicitly identified grant objective (and almost always an implicit objective) and, therefore, a target to be measured by program evaluators. Collaboration among the various organizations and individuals involved in these multi-faceted approaches is viewed as essential for success (Center for Mental Health in Schools, 2003; Gajda, 2004; Riggins, 2004). Typically, it is hoped that collaborations formed for the purposes of implementing the grants will remain after the funding ceases. Collaboration is seen as a prerequisite for sustainability of interagency programs (Hogue, 1993; Perkins, 2002; Peterson, 1991), particularly for those programs initially created with the support of time-limited funding sources. From the perspective of evaluators, however, assessing collaboration is often difficult. Models of collaboration among agencies, groups, and community stakeholders are notoriously difficult to translate into valid and reliable instruments which can measure meaningful change in the level and pattern of collaboration. This paper reviews the existing models of collaboration, presents an instrument for its assessment consistent with common characteristics across models, examines reliability and validity evidence for the instrument, and suggests a method of graphical display to represent the sometimes elusive nature of collaboration. Models of Collaboration