Evaluations of special education interventions are different from research on medical interventions. Educational evaluations would be more useful if they applied an actuarial, or probability approach. Information should be collected over time and using various measures, thus increasing the available data. (GDC)