The core work of educational administration is school improvement, especially improved school outcomes. The leader must know what to do, why some interventions work better than others, and how to implement them. Longitudinal experimental research, meta-analyses, and evaluations have definitely shown that small classes, at least in K-3, provide an array of positive schooling outcomes. Part of the "what-to-do" dimension. Drawing on 15 years of analyzing class-size effects, this synthesis situates small-class outcomes in research and theories to explain why small classes provide initial, increasing, and long-term benefits. Supporting research includes Head Start, Perry Pre-School, and the Abecedarian Projects. Using Brookfield's (1993) definition of professional "impostorship," impostorship in educational administration will not sell in the face of increasing public demands, scrutiny, and alternatives to public schooling. A recent survey showed that fewer than 30 percent of educational administrators (average of 10 years' experience) could list even one research-driven education improvement effort they learned in educational administration preparation programs. Without clearly knowing what, why, and how of education, educational administration cannot improve school, an ever-present task. (Author)