This paper explores the functioning of the low-inference observation system used as part of the Tennessee Career Ladder evaluation system, which provides teachers with the opportunity to qualify for additional pay on the basis of outstanding performance. The analysis of the functioning of this form is intended to provide the basis for the ongoing improvement of Career Ladder evaluations and to increase understanding of the role of low-inference observation in the evaluation of teachers. The data reported are taken from Career Ladder data from the 1984-85 evaluation, but the metric used is not the same as that used in the actual evaluations. A number of changes in instrumentation and scoring methods were made after the 1984-85 evaluations. Data are based on a random sample of 275 teachers statewide and on actual evaluations of 3,132 teachers in 1984-85. The following conclusions were made: (1) the choice of aggregation level determines the nature of the analytic outcome; (2) the choice of scoring method exerts a powerful influence; and (3) the potential of pattern scoring is matched by the difficulty of developing meaningful scoring formulae. An appendix includes the observation form. (JAZ)