Abstract
Self-reports of behavior are the most commonly used method for collecting outcome data in health education programs, but they are subject to a variety of sources of measurement error. Problems of measurement in some recently reported education program evaluations are identified, and a model is proposed to identify systematically the major influences for accuracy of self-reported measures of behavior. Several methodologic studies on the skill aspect of self-report form completion are briefly described, and eight steps to promote more accurate self-report of health behaviors are proposed.