''The idea is to integrate the research act into the educational setting so that it can play a direct and immediate role in improving the practice,'' said Professor Kemmis, who is also co-director of the Centre for Education and Change at Deakin.

He suggested teachers treat their own educational ideas and theories, work practices, and settings, as objects for analysis. ''With careful reflection, teachers may uncover theoretical ideas or assumptions which may turn out to be unjustified,'' said Professor Kemmis.

For example, some teachers were too rigid in their assumptions about the nature of their students' abilities, he said.

In other cases, teachers might find some practices, acquired through habit or tradition, irrelevant or no longer useful.

Professor Kemmis said action research followed a spiral of cycles of reconnaissance, planning action, enacting and observing the planned action, reflecting on the implementation of the plan in the light of evidence collected during the implementation, replanning, further action and further observations.

''These activities are aimed at improving the practice, understanding and situation, and involving as far as possible, those affected by the action research process,'' said the professor.

Co-ordination and involvement of teachers and others in data-collecting, analysis and critique would not only help researchers in finding new ways of thinking, practising and structuring educational settings, but also create an immediate sense of responsibility for the improvement of the practice, he said.

''Participation in action research is thus a form of professional development.''