Four types of science education research are defined: empirical, philosophical, policy, and developmental studies; problems and approaches related to each type are analyzed. Problem definition is discussed in terms of theoretical and practical problems. Arguments are presented for and against requiring doctoral students to define their own problems. Use of levels of significance in empirical research is discussed and a distinction drawn between statistical significance and decision making significance. The need for philosophical studies which analyze assumptions and the consequences of actions is stressed. Historical and comparative studies are advocated as aids to making policy decisions. A pattern for developmental research is outlined. The responsibility of the researcher to suggest the implications of all types of research is discussed. (EB)