This paper reports the analysis of the results from a pilot effort to create and use a battery of instruments based on INTASC principles indicators of teacher dispositions. The original conception of the battery was designed on the taxonomy of increasing levels of inference. This means that the intent to measure included multiple instruments in order that confidence in the assessment was greater than typically found in cognitive tests. For example, if a teacher passes a teacher certification test once by correctly answering pedagogy and content questions, there is an assumption that the teacher candidate knows and can do the material on the test. Obviously, guesses, cheating, and simply knowing (but not willingness to do) are all threats to the confidence that the test taker possesses the knowledge and skills referenced. In the areas of affective assessment there is always the possibility that someone would self-report what was "expected," but not what they really believed or behaved, so confidence that the assessment has value as a true score of the person tested is important. There are several ways to gain confidence in affective assessment. One is to design instruments that "pull" beliefs and values from the test takers where faking is difficult and detectable. This is achieved by making questions in formats that have dissonance such as agree-disagree statements, questions which require specificity so that fake answers are spotted, and observations of behaviour consistent with beliefs. Multiple measures with different item types are the best way to gain the confidence we seek. (Contains 5 tables.)