We seek to understand the process by which a school incorporates or enacts an externally developed reform design. An externally developed school reform design is a model for school improvement that is developed by an outside design team. This team generally conceives the reform design; develops the principles, implementation strategy, and materials that accompany the reform; and sometimes provides training and supports that enable local schools to prepare educators to implement the reform. When implementation of a tested prototype program or design expands to many schools, the process is known as replication or, in the current educational reform literature, scaling up (Elmore, 1996; Stringfield & Datnow, 1998). Scaling up has proven to be a vexing and seldom successful endeavor (Elmore, 1996). We argue that this is due to a lack of understanding of the co-constructed nature of the implementation process.
Studies that treat the implementation process as uni-directional, technical, mechanical, and rational (Carlson, 1965; Havelock, 1969) do not fully capture how educational innovations play out as social, negotiated features of school life. Organizational models of school improvement that developed in reaction to these technical-rational models also do not suffice for understanding school reform implementation (see, e.g., Fullan, 1991; Louis, 1994). Because their focus is on school-level strategies for self-renewal and improvement, organizational models downplay the actions that initiated the reform and the governmental, community, and district actions that occurred away from the school before it attempted rejuvenation and renewal. Neither technical-rational nor organizational development models help us fully understand educational implementation, which we believe involves a dynamic relationship among structural constraints, the culture of the school, and people's actions in many interlocking sites or settings.
Our research builds upon work in the sociocultural tradition that has helped shape the Center for Research on Education, Diversity & Excellence (CREDE), especially Rogoff (1995) and Tharp (1997, p. 12), who identify personal, interpersonal, and community "levels" or "planes" of interaction; and McLaughlin and Talbert (1993), who depict organizations as successively contextualized layers. We extend this work by explicitly calling attention to the political and economic conditions that enable possibilities and impose constraints on education in general and on school reform in particular. We also try to avoid privileging any one context in our discussion of educational implementation by showing the reciprocal relations among the social contexts in the policy chain.
We believe that formulating the reform implementation process as a "conditional matrix" coupled with qualitative research is helpful in making sense of the complex and often messy process of school reform. To illustrate our formulation, we report on two CREDE projects: a study of the implementation of six reform efforts in one Sunbelt school district1 and a study of the implementation of the Advancement Via Individual Determination (AVID) untracking program nationwide.2 We hope that our formulation will be helpful to others studying the school reform process.
Before we trace the implementation of reform efforts in the schools involved in our two studies, we present the assumptions and premises that guide our research.