What Good are Statistics that Don ’ t Generalize ?

Quantitative and qualitative inquiry are sometimes portrayed as distinct and incompatible paradigms for research in education. Approaches to combining qualitative and quantitative research typically " integrate " the two methods by letting them co-exist independently within a single research study. Here we describe intra-sample statistical analysis (ISSA) as a general technique for using quantitative tools to support qualitative inquiry so as to simultaneously provide warrants from qualitative and quantitative traditions. In certain circumstances ISSA makes it possible to relax the requirement that individual participants be treated as the unit of analysis in statistical models, and thus provides justification for coding qualitative observations and drawing statistically based conclusions about observations in a qualitative context. We developed ISSA and describe it here both because it can be used as a tool for qualitative research, and because it illuminates the relationship between method and interpretation in the research traditions that it bridges. In this article, we (a) summarize key features of qualitative and quantitative research relevant to ISSA; (b) describe ISSA as an analytical technique; (c) discuss the quantitative and qualitative justification for ISSA and the nature of the conclusions that can be drawn based on it; and (d) explore the more general implications of ISSA for qualitative and quantitative inquiry. T hough written about a border skirmish in the " Great Game " of colonial power at the end of the 19th century, these lines from Kipling's " The Ballad of East and West " aptly describe the current schism in education research method-ologies. Some have declared the end of the " paradigm wars " between qualitative and quantitative methods (Tashakkori & Teddlie, 1998, p. 1), but debates about the comparative validity and utility of research traditions 1 continue, not least in recent critical reports about the nature of education research (e.g., National Research Council, 2002). The persistence of methodological divisions is hardly surprising. Although qualitative research and quantitative research are really collections of loosely related methods and associated techniques rather than natural categories, substantive differences do exist between them. These differences result , ultimately, from differing perspectives on the nature of fundamental and problematic concepts such as " truth " and " understanding , " on the kinds of claims that are worth making, and on the kinds of warrants required in scholarly inquiry. The lines that follow Kipling's oft-quoted description of worldviews in conflict are less well known, but are …