Models used for evaluating teacher education programs are described, particularly the learner referencing educational evaluation model. This model involves ten stages: (1) acquisition of possible teacher training program standards; (2) verification of those standards; (3) generation of specific learner objectives; (4) processing of learner referenced data to determine specific objectives; (5) external validation of university standards by classroom applicability; (6) comparison of teacher training program standards with learner performance measures; (7) generation of strategies for eliminating discrepancies between training program standards and learner performance; (8) consideration of adjustment strategies for the training program; (9) incorporation of these adjustments; and (10) determination of the effectiveness of these adjustments. The limitations of this model, and its initial implementation are described. Graduate-level student teachers from the University of Northern Colorado developed individualized education programs for their learning disabled and speech handicapped students. Data were collected regarding the attainment of these behavioral objectives. Discrepancies between the standards set by the teacher education curriculum and the actual classroom performance were studied by the university as part of the effort to improve its programs. The student data and some of the student teachers' and cooperating teachers' comments are included. (GDC)