An instructor in a junior high school English program which involved teaching approximately 160 eighth graders on a daily basis found it almost impossible to assess students fairly by taking individual differences into account when grading. Her solution was to invite students to participate in the construction, performance, and evaluation of their own learning. Eventually the students were trained to both formatively and summatively evaluate the writing of their peers through the use of focused interactive dialogue, and it was found that this experience with student evaluation of peers' compositions established the validity and usefulness of peer ratings and grading; it also resulted in a number of reflections about the nature of grading and student assessment. Collaborative research by three male and three female teachers studied student perspectives on evaluation measures, beginning by examining the teachers' grading and evaluation practices. Five were currently teaching secondary school, and 4 taught 130 to 158 students each day. All six teachers indicated that they gave grades only because the educational system required it. Two teachers in alternative schools did not give grades as such, but offered summary judgments of student progress through passing or failing. One teacher offered formative evaluation but did not grade her students; the students graded themselves through discussion with the teacher. Reflections on these teachers' approaches are presented first in the context of the traditional mimetic teaching model in which knowable truths are taught and their acquisition is monitored and measured by evaluation, and then with respect to a "maieutic" model of exploring, integrating, and transcending (creating personal meaning). The collaborative research of this group of teachers illustrates many of the problems associated with student evaluation in the mimetic tradition. A 42-item list of references and the original evaluation and grading survey are included. (SLD)