A model is proposed for the clinical synthesis of data from psychological tests of persons with visual impairments. The model integrates the concepts of the ideal type and Bayesian probability and compares actual test scores with ideal scores through use of a pattern similarity coefficient. A pilot study with Business Enterprise Program operators illustrates model application. (Author/DB)