This study of student teachers in high school mathematics and science classrooms compares their teaching behaviors and their scores from three pencil-and-paper tests. Data analysis indicates a relative independence between the ratings of teacher classroom behavior and scores on the pencil-and-paper tests. There was little consistent relationship between how well student teachers could judge what they ought to do in a classroom situation and what they actually did. Apparently, personality dimensions, as measured by the Myers-Briggs, tell little about what to expect from a person in the classroom. (Author/NK)