Formative and summative evaluative methodologies are distinguished. The need for internal, external, statistical, and construct validity of all correctional methodologies is emphasized. Foundations of behavioral research and methodology in the social sciences are reviewed. Experimenter effects are considered to have a profound influence on treatment outcome. General impediments to the effective use of psychological diagnosis in criminology are outlined, and proposals are offered for new research designs. The discussions are addressed to sociologists, psychologists, educators, social workers, and criminologists. Footnotes and references are included.