This article describes an exercise that teaches students about methodological issues concerned with making reliable observations of behavior. After learning Ekman's (1972) Facial Affect Scoring Technique from a microcomputer program simulating expressions of emotion, students recorded the facial expression, gender, and age of people in natural settings, computed interobserver agreement scores, and submitted a laboratory report. This exercise generated much discussion about research methods, transferred skills from the classroom to a research setting, and illustrated our view that the microcomputer is a new tool that supplements, but does not replace, existing instructional techniques.