Urban ethnography and research integrity Empirical and theoretical dimensions

Problems of data reliability and validity continue to plague the social and behavioral sciences. Enhanced reliability of quantitative data has been achieved by such means as greater attention to instrument construction and testing, improvements in survey technology and sampling of targeted populations, and careful operationalization of theoretical ideas. The integrity of much quantitative research is compromised, however, when quantitative measures lack validity, especially when properties of groups and other collectivities are measured by responses obtained from individuals. Ethnographic and other forms of observational data offer important correctives to this problem. We illustrate the interplay of quantitative and qualitative research methods by examining studies extending ideas from Elijah Anderson's Code of the Street to 'neighborhood codes of violence', the violence-avoiding function of street codes, and special problems of evaluation research as they apply to programs designed to control street gangs.