I hardly know where to begin in responding to Ralph Scott's farfetched analogies, his misguided notions of how one gets data for analysis and reanalysis, his non sequiturs, his own unpublished “fugitive” references, and more. Only a misguided sense of fairness and interest in reporting both sides of controversial issues could have prompted the publication of a piece so full of factual errors.
Anyone familiar with my work is aware of the utter foolishness of Scott's claim that it has been impossible to examine the empirical basis of my conclusions about tracking's harm to low-income and minority youth. My conclusions are based on a number of widely available, published studies, and most recently, on my Multiplying Inequalities study (1990). That study analyzes data from a federally funded National Science Foundation survey of 7,200 teachers and principals in 1,200 schools (1985–86).
My research methods are detailed in the RAND report that I sent to Scott (at my expense), even though it and the rest of my work are available at any major library. There he will also find most of the studies in the more than 300 references that I sent him. Many of them are also analyses of federally funded, publicly available databases.
Ironically, even worse than Scott's unfounded personal attacks is his own use of the bad social science that he carps about. He makes unsubstantiated, sweeping, and patently false assertions when he says (1) that other industrialized nations employ more and earlier tracking than the United States, and (2) that the SAT decline in the early 1980s was caused by “significant applications of untracking.” He illogically declares that GM's granting subcontracts to Japanese firms is an example of the consequences of untracking, and he uncritically accepts opinions expressed in The Economist because they match his own experience and ideology (“As someone who attempted to effectively teach..., it is difficult for me to fault the viewpoint that encourages tracking”).
There are certainly discussions to be had about standards of evidence in social science and the use of research to inform public policy. Much more must be said about issues of tracking, ability grouping, poverty, minority status, and school reform. Unfortunately, Scott's piece does nothing to advance any of these discussions.
