This study continues an earlier project (ED010369) to determine if student achievement and attitude toward school subjects (1) can be improved by increasing feedback to teachers concerning pupil and/or trained observer reaction to their teaching, (2) correlate significantly with attitude of teachers toward their pupils, (3) can be improved to a greater degree by face-to-face feedback to teachers than by standardized feedback via mail. A pre- and posttest 2 by 4 factorial design was used with the sample of 80 teachers and 1,912 sixth grade students in 7 Pennsylvania school districts. Statistical analysis of the data using class means revealed no significant differences in achievement or attitude, but analysis using individual scores indicated significant differences favoring written feedback, particularly that of pupil reactions to teaching. Teacher opinionnaires indicated that benefits derived from both pupil ratings and classroom interaction analysis warranted their wide use in teacher inservice programs. Overall implications are that it is necessary to provide frequent and intensive help to teachers in order to change teacher behavior or effectiveness. Included are a 48-item bibliography; a 17-page manual, "Interaction Analysis: A Program for Interpretation of Matrices;" the Hayes Pupil-Teacher Reaction Scale; the teacher's Feedback Study Opinionnaire; and 48 tables of statistical data and analysis. (JS)