Abstract
In spite of claims to the contrary, affective factors related to teaching and learning can be measured. Recent development of affective scales in reading attest to this, yet caution must be observed that as new-scales are developed, they are adequately validated so that the research emanating from their use can be regarded with a high degree of credibility. In that light, those who construct affective scales must consider more than content validity; they must also investigate construct validity. This includes evidence in four categories: reliability, convergent validity, discriminant validity, and sensitivity to treatment. Specific references are made to investigations of the validity of the Vaughan Attitude Scale which measures teachers' attitudes toward teaching reading in content area classrooms.