This paper examines educational research on student development and discusses some of the reasons why this research has had so little impact on actual educational innovation. These include: (1) failure to anchor research within a more general model of development; (2) lack of in-depth descriptive and phenomenological studies of students that would identify what types of changes occur more prominently than others in what types of colleges; (3) too much emphasis on objective, highly focused tests of the developmental process; and (4) absence of adequate methods to assess the specific types of educational programs that produce different types of psychological effects of development. The paper then reviews the principal features of one developmental model based on five interrelated dimensions of maturing: symbolization, allocentricism, integration, stability, and autonomy, and explores the model's implications. (AF)