The Continuing Hope: A Rejoinder

Abstract

The most persuasive critique of my conclusion that variations in school resources are not related systematically to students' would be evidence to the contrary. But none is offered by Spencer and Wiley, and none is to be found. The main substantive argument in their article relates to the possibility that schools or teachers pursue different goals. They argue that measures of that do not reflect differences in goals might be misleading. School resources may be related to the achievement of these goals even though unrelated to performance as measured in the studies that my article summarizes. Their discussion is quite unspecific about the nature of these different goals, and it contains no suggestions about how research should be fashioned to test their idea. Nor do they present any evidence that differences in goals are truly important in understanding school resource use. The only explicit example of goal differences that they provide is new math. However, this example does no more than illustrate that schools indeed do a poor job of decision-making, a subject that is the heart of my analysis. From a public policy view, it is probably unwise for society to accept unquestioningly the goals of any school or any teacher. Suppose a school system chooses a very specific but limited goal (say, that all students should know that the first word in a sentence should be capitalized). Without evidence of in other dimensions, we probably would not applaud even perfect attainment of such a goal. Most standardized achievement tests, on the other hand, are designed to measure such things as basic reading or math skills-skills whose value is widely accepted by society. In any case, many of the analyses reported in my article do consider the possibility that must be measured in more than one dimension. Thirty eight of the 130 studies covered in my article consider jointly determined outcomes, using the appropriate statistical techniques. Looking at these studies alone,