In order to address the complex question of school reform in multiethnic cities, we necessarily drew on multiple data collection and analysis methods. No one method is privileged in the analysis; the different methods described here are all seen as important but partial approaches to the questions we are interested in. Bringing together the information with diverse methodologies provides a more nuanced understanding of the different dimensions and meanings of school reform issues.
It is useful to keep in mind the broader purposes of the National Science Foundation (NSF) study. The focus was on the context of public education, rather than on an analysis of the classroom dynamics or an evaluation of particular programs. The conceptual focus was on the political coalitions and dynamics in which education policymaking is embedded and the different types of civic arrangements associated with various reform efforts. School systems are seen as one part of an interrelated network of political institutions situated at the local level, but not necessarily controlled entirely by local officials. Schools’ capacity to make innovative choices is contingent on these contextual features and collaborative political arrangements. The overall project goals, therefore, included mapping the patterns of alliances and civic dynamics providing the context in which school policies are formed; assessing the range and variation in education and other human capital investment initiatives in each city; and identifying the coalitions, decision processes, and ideas associated with different school reform initiatives.
Eleven cities were included in the NSF project: Atlanta, Baltimore, Boston, Denver, Detroit, Houston, Los Angeles, Pittsburgh, St. Louis, San Francisco, and Washington, DC. All have substantial minority populations and large numbers of students from poverty households. The field research began in May 1993 and extended through December 1994. The initial hypothesis postulated that more systemic school reforms were more likely in cities with more civic capacity. The dependent variable in the larger study measured human capital investments in terms of the aggregate effort (tapping wealth and school expenditure measures), the scope of the effort (the number and types of programs serving disadvantaged students), and the extent to which these efforts constituted systemic reform (rather than incremental tinkering). For each dimension, there were attempts to determine whether the initiatives were sustained, whether the efforts were real or symbolic, and whether they were targeted to the disadvantaged in the schools. Data for these measures came from aggregate statistical sources as well as from interviews in each city.