The general purpose of this study was to compare the differences in middle school students' mathematics achievement, their changes in attitude towards mathematics, and their attitude towards evaluation when evaluated with two different measurement strategies. The primary purpose of the study was to compare aspects of criterion-referenced and norm-referenced evaluation within selected sixth and seventh grade mathematics classes at the University of Northern Colorado Laboratory School. The design for this investigation was quasi-experimental nonequivalent control group design. Ninety-five students were assessed in regard to mathematics achievement, attitude towards mathematics, and attitude towards elevation at the beginning and again at the termination of the 12-week trimester. Overall, students obtained higher achievement scores when evaluated using a criterion-referenced method keyed to the specific performance objectives of an individualized instructional program. Students evaluated by criterion-referenced methods demonstrated significantly more positive attitudes towards the subject than did those middle school students evaluated by norm-referenced methods. Middle school students indicated no preference for either the criterion-referenced or the norm-referenced methods of evaluation. (Author/PN)