We appreciate the thoughtful critique of our article and the opportunity for this dialogue in Exceptional Children about some of the challenging issues in education. We agree with Soodak and Podell that major reforms of are needed and that they should occur within the larger context of educational change. It is our assumption that all categorical programs (including education, Chapter 1, migrant education, and others) face serious reexamination in the near future. It is our hope that these programs will be reformed in a coherent way with each other and with general education. Even in favored suburban schools--and wherever there are general changes, such as the move toward school-based management--there is a need for new approaches in categorical programs. The school is the most fundamental unit for managing changes in education; and that is why we organize 20/20 analysis around the individual school. Our article discussed only the first phase of the 20/20 procedure. Immediately after the initial data analysis, attention shifts to a second phase, which is concerned with improvement of instruction for individuals and for the class or program as a whole. The idea is to move directly and positively toward program improvements and, as much as possible, to avoid tendencies to excuse poor progress by applying negative labels to children and then setting these children aside in special programs. In this response, we discuss several of the specific issues raised in the critique: inequity (to which we join the issue of funding), single-dimension identification, and simplicity. INEQUITY A key point in the critique concems the creation or maintenance of inequities among schools, and thus for students, if 20/20 procedures are used. This point concerns funding, a topic that deserves much attention but that we did not discuss. We do not propose that categorical funds be distributed uniformly across schools (20% or 40% to each). We now have 20/20 data on dozens of schools; and the differences among them in achievement levels are enormous, much as shown in the critique comparing an inner-city school and a suburban school. Obviously, some schools will need more resoumes than others. We favor an approach to funding suggested by Nicholas Hobbs (1975), in which programmatic units serve as triggers for funds. As education, Chapter 1, and other categorical programs are reformed (and linked with changed delivery systems in health and social services), we will need to define programmatic units and how they are funded. Some of the programs may be directed to prevention of problems and thus not involve identifying particular students. Perhaps extra teachers and parent education will be programmatic units. In any case, we need to get away from the bounty hunt mentality, in which more labeled students beget added dollars. Present funding systems provide too little in the way of programmatic traces for categorical dollars. Indeed, in today's cash-short school environments, one might guess that school districts are eagerly seeking categorical dollars, but reducing their flow to categorical programs--all without clear programmatic traces. Clearly, funding is an equityrelated topic and deserves much discussion. We simply did not deal with it in the initial article. SINGLE-DIMENSION IDENTIFICATION We note that using general reading ability as the key dimension in 20/20 analysis will result in the identification of most students now in (more than 75%). We also recognize that reading is an important and continuous dimension of achievement, that it is among the variables most predictlye of school dropout and other problems of students, and that it does not assume a typology (no direct labels). …