A paper and pencil instrument was developed consisting of statements about teaching practices appropriate for students of ages 10 through 14. The statements are to be accepted or rejected by the respondent teacher. This paper reports the procedures and results of a third field test and the design of a fourth field test. The third field test instrument consisted of 244 items. It was administered to 320 teachers in eight schools. Two hundred of the items underwent statistical treatment: item analysis and factor analysis. The next step was to identify which of the 200 items correlated with data from classroom observations. Four observation instruments were used. The raw scores from the observations, combined into sets and converted into ratios were then correlated with the paper and pencil item responses of the 40 teachers. Five statistical criteria were computed by which to check the strength of each paper and pencil item. Through these criteria 97 items were eliminated and 103 emerged. Of those that emerged: (1) Four-fifths of the items correlated both statistically and logically with one or more sets of observation data; (2) Forty-five of the items loaded in factors; (3) Twelve neither correlate with observation data nor load into factors but were retained because they discriminate well between high scorers and low scorers. When the instrument is fully developed it will be used as one part of a pre-training inventory procedure by which teacher trainees can obtain an estimate of their general competencies and a diagnosis as which to need development. Finally, systematic observation of classroom processes discriminates between teachers in many useful categories. (CK)