The reliability of the Writing Skills Assessment Test (WSAT) given to all entering freshmen at the City University of New York (CUNY) is the subject of this monograph. The monograph states that although the WSAT is scored according to validated and widely-used holistic procedures, some CUNY faculty had expressed concern about aspects of the WSAT's question wording. The study reported on here experimentally manipulated two aspects of the WSAT questions in an attempt to determine whether variations would increase the test's potential for eliciting students' minimal writing competencies. The aspects manipulated were (1) cognitive demand levels--the degree of information and the number of strategies which an essay question provides; and (2) experiential demand levels--the degree of personal experience for which an essay question asks. The results showed that neither of the changes had any significant effects on students' writing performance; these findings provide empirical construct validation of the WSAT. Appendices to this monograph contain a sample WSAT form, the evaluation scale used with the WSAT, sample (annotated) student essays representing each point on the scale, the essay questions used in this study, and inter-coder reliability data. (CMG)