Component-building research on practical issues of instructional design can make a substantial contribution to research-based school improvement. Component-building research consists of rigorous within-school experiments that test various components of instruction for effects on student achievement. Research would address practical issues that teachers and principals face each year--for example, should homework be given every day, should films be used more or less often, should a school use a reward-for-effort or continuous progress grading system, should students be regrouped for mathematics instruction, and so on. A large number of such rigorous studies conducted by teachers, principals, and researchers could produce a science of instructional design. Complete programs of effective instruction could be developed by combining components that are proved to increase student achievement. (Author/JD)