The problems stated at the outset were: to discover whether information provided by confidence testing would result in a more accurate assessment of student knowledge so more personal individualized instruction could be presented, and whether sufficient precise data could be attained to be of significant aid in curriculum planning. Analysis has shown that much qualitative and quantitative information is provided when employing the methods and materials associated with conficence testing. Predictive validity appears to be positive, but not significantly high enough to forecast a student's performance on practice teaching lessons or his final standing in the class. The assessment of student knowledge tends to be more accurate when using confidence tests. Reliability of test instruments seem to be improved when using confidence tests. Confidence tests provide more information about the students than do choice tests. Better item analysis data is provided with confidence tests than with the standard choice tests. (Author)