"Action research" is involved in any evaluation of an initiative where the researcher also forms part of the team responsible for the initiative, or where the researcher actively seeks to guide and influence the initiative while it is underway. Action research is not currently characteristic of research in the prison system. Treatment projects would lend themselves to action research, when an initiative is intended to provide treatment of some kind to specific inmates, or to help staff in a particular situation. Other types of correctional projects suitable for action research are assessments or prison management initiatives. Often the "action" part of the research will be for the researcher to help the team that is implementing organizational change to develop its aims in such a way that they can be measured and the progress of the initiative monitored. The stages of an action research project are to define the initiative/project, to analyze the current environment before change occurs, to design the evaluation, to implement the project and the evaluation, and to complete the evaluation after the official termination of the project. The major pitfalls in action research are in the fourth and fifth stages: implementation and analysis/report writing. In the implementation stage, the danger is that the action researcher focuses so much on getting the project underway that the evaluation schedule is neglected. A pitfall at the analysis stage is that the researcher becomes so committed to the "success" of the project that objectivity is lost in the evaluation. 7 references