The purpose of this study, conducted in 1994, was to solicit feedback from multiple appraisers (students, administrator and teacher) on the teacher's performance and compare the results to see if there is dissonance (or significant differences) in the three groups of raters. This study was conducted in three school districts, two in the United States and one Department of Defense district in Germany. A total of 6,771 students, 15 principals and 182 teachers gave feedback on teacher performance;The feedback was obtained using three questionnaires. The student questionnaire was a pre-existing instrument that had been used in previous studies and had proven to have validity, reliability and legally discriminating power. For this research, under the auspices of the School Improvement Model at Iowa State University, two companion instruments were developed. The teacher, self-feedback and the administrator questionnaires were designed to parallel the student questionnaire and elicit information on the quality of the teacher's performance;There were four groupings of questionnaires for all three rater groups: lower elementary (K-2), upper elementary (3-5), middle school (6-8) and high school (9-12);All three rater groups completed the questionnaires rating the teacher using a five-point, Likert-type scale. The data from these questionnaires were analyzed using the SPSS statistical software program created for two previous studies at Iowa State University Computational Center;The data were then analyzed to determine which rater group, students, principals or teachers rated the teacher with more leniency or severity;The findings, consistent across all three districts and all three rater groups, revealed the following: (1) The modifications of the survey forms for rating teachers were successful. The "I do" form for teachers and the "This teacher does" form for principals created for this investigation worked well and elicited parallel responses which allowed comparison of data from the three rater groups. (2) Students, at all four levels, consistently rated the teacher more severely than principals and teachers. (3) Principals, at all four levels, consistently rated the teacher more leniently than students. (At the K-2 level, principals were the most lenient raters.) (4) Teachers, at all four levels, consistently self-rated more leniently than students.