The papers presented at this conference review current research on effective educational practices for low-achieving students and identify those that might be applied to programs funded under Chapter 1 of the Education Consolidation and Improvement Act of 1981. A function assumed by several conference participants was to act as reactor to papers written for a particular session. This report reviews and discusses three papers prepared for the session on program and staffing structures and dealing respectively with summer programs, instructional setting, and student-to-instructor ratio. The paper maintains that, once program auditing was instituted by the federal government, the instructional settings began to be chosen by local education agencies (LEAs) for Chapter 1 programs mainly to escape "audit exceptions" and that they tended to stick with the pullout programs (as opposed to supportive service projects) because they were "safe and clear." Later changes necessitated that children remain in class, and once again the delivery system was altered in response to public outcries for change,rather than on the basis of sound pedagogical theory. As a result, these changes generated no significant gains in achievement. Teacher training, improved data collection, and careful, controlled, research/evaluation of the various alternative approaches to providing Chapter 1 services should be undertaken. Therefore it is proposed that school districts be permitted to apply for a waiver from Federal and state requirements to conduct experimental programs in Chapter 1. A review procedure should be established in the Department of Education to assure that the model developed is "audit exception free." LEAs should be allowed to develop innovative approaches to solving problems. Programs of this type should be given a three-year experimental status, have a tight evaluation design, a clear audit trail, and be reviewed at the end of each year for continuation approval. When school districts are given the opportunity to develop creative program designs, more effective Chapter 1 programs should result. A list of references is included. (PS)