Article reanalyzes and expands upon data from Tennessee's Project STAR which examined the effects of class size reduction on student achievement in the primary grades. It describes six competing theories of class size impact on achievement and test performance, settling on the student group/modeling interpretation of study data. (SM)