Methods, interpretations, and different views of aspirations

Abstract

Ellen Brantlinger's paper in the last issue of lnterchange--"What Low-Income Parents Want from Schools" raises the issue of methodological scholarship and class bias in interpretations of aspirations and schooling. An article penned by Diana Crane in the late 1960s---"Fashion in Science: Does It Exist?"---comes to mind as a useful way to start the discussion. Crane's title, I believe, came closer to truth than did her analysis: fashion, false starts, data framed by their own generative categories are part and parcel of inquiry. Science as method is built upon scepticism, provisionality, tentativeness, specification of conditionalities, and, above all, uncertainty. It is only within these constraints that logic, empiricism, parsimony of findings, and the like give up provisional "truths." But method can become a kind of cult itself which, ultimately, undermines understanding. In the social sciences, we witness an irrational belief in the power of regression as "scientific," as if numbers themselves could be treated and understood prima facie apart from the social context which generates them. They are treated de facto as if correlations are not every bit as subject to interpretive processes as qualitative approaches. Such presentations are technically, of course, "mis-uses" of method, but they are indicative of a belief that one kind of method will generate a more valid type of knowledge than another. Such a belief is fashion. An inquiry which is authentically open must go beyond fashions and draw insights where it can. This is helpful to recall in commenting on Brantlinger's article. First, we can anticipate how others will criticize her work: part of methodological scholarship involves comparing ideal design with actual practice. In Brantlinger's work, the data contain a variety of potential limits to their reliability and validity. The data are non-parametric and small; the sample is atypical for its target population (most low-income parents live in large cities or rural areas, not middle-sized cities); lower income families are not over 90 percent white; and so on. Other criticisms may be made of the qualitative methods: questions of inter-observer reliability, inter-subjective agreement on procedures for interviewing, the degree of systemizing the content analysis, checks for split-half validity, issues of countervailing interpretation can all be brought into question. Presented only on the basis of data gathered by the researcher(s), the case for either external reliability or validity is, at best, tenuous. Raising such issues is rightly the critic's role. But it is not the critic's only role. The critic's other role is to glean insights from what we do know that advance our knowledge. This aspect of reviewing is often forgotten in favor of dealing solely with the method. But such reviews are frequently little more than paradigm disputes. In