The purposes of this paper are to (1) discuss general issues related to a newly proposed methodology for item content and construct validation and (2) apply the new method in the context of developing a scale for measuring attitudes toward black people. Latent partition analysis was proposed and applied for the study of item content validity using partitioned data from 36 college presidents. Factor analysis was used to assess construct validity using item response data from 212 college students. A comparison between the item response factors and the LPA categories indicated substantial agreement. Limitations and virtues of the present methodology are discussed in relation to certain other methods for content and construct validation. (Author)