These models try to identify theoretical dimensions underlying correctional agencies' operations and to clarify and order concepts. The single-dimension models specify a single goal dimension along which particular agencies, practices, or staff are expected to range. The extremes represent cultural or political orientations. Packer's crime control and due process models of the criminal process are the best-known example in the criminal justice field. Discussions on prisons have focused on custody versus treatment or punishment and rehabilitation. Dembo's model is the best-known model in correctional field work, particularly in parole. These single-dimension models have led to an emphasis on the ineffectiveness, inefficiency, and sometimes the immorality of combining control and service roles in the same agency or the same set of programs. Some recent studies have raised doubts about the accuracy of analysis based on the single-dimension models. In the 1960's, more complex models emerged. These regarded the previous conflicting goals as separate dimensions and described both caseworker types and agency or worker orientations. However, the two-dimensional models have problems as well. All the models overlook many factors, but each includes some crucial elements of supervision. The models have identified several different levels of concern by focusing on the goals, the procedures, and the people working in agencies. Corresponding to the shift to more complex models has been a shift in emphasis from acceptance of conflict to resolution of conflict between roles and agencies. Further research on the variables involved in the models and on the relationships among the variables is needed. Forty-eight references are listed.