This paper discusses the integration of positivist single-subject design with interpretive, or naturalist, qualitative methods in the development of effective interventions for rural special education students. The resulting research/intervention design includes: (1) extending the normal baseline period from 3 to 10 days thus allowing the researcher to question all relevant peers, adults, and the student him/herself in order to investigate multiple realities, triangulate the data, and ferret out a research design and an initial independent variable; (2) applying the independent variable in the classroom setting while continuing to monitor multiple realities in order to fine-tune the design and discover other independent variables; (3) returning to the baseline phase without the application of the independent variable; and (4) applying a new independent variable based upon information gathered in the first three phases and comparing the effects of the first and second independent variables. The synthesis of positivist and interpretive methods results in the following theoretical assumptions: (1) there are multiple realities; however, the reality of the targeted research participant is of primary concern; (2) the research participant, teacher, and researcher continually interact, calling for an understanding of differences in realities; (3) research seeks to generalize theoretically, not statistically; (4) causality of realities may demonstrate functional relationships; and (5) the values of everyone involved in the research must be defined to control bias. This method allows researchers to connect student behavior with internal states and subsequently attempt to affect internal states through behavior intervention. (LP)