Ethnographic Data Collection in Evaluation Research

quantitative designs (e.g., Eisner, 1979; Filstead, 1979) have contributed to the increasingly common phenomenon of incorporating an ethnographic component into evaluation research. The systems-analysis approach (Rivlin, 1971), commonly used in economics and industry and which has dominated federally funded educational evaluation, emphasizes measurement of variables that are scaled or quantified easily and that generate highly reliable data. In many cases, this approach has been inappropriate to the phenomena under investigation and has produced data of uneven and questionable validity (Guttentag, 1977; Helfgot, 1974; House, 1979; LeCompte, 1970). The results have failed to satisfy those involved in programs and their evaluations, to explain success or