Several analyses of the construct validity of the fourth-grade, eighth-grade, and commencement-level English and Mathematics examinations of New York state were performed. The analyses present construct and differential construct elaboration both across tests and within tests. Results show strong relationships among different question types, open-ended and multiple choice, within the same tests and weaker relationships for similar types of questions in diferent tests. These findings indicate that the tests are much more sensitive to skills they are designed to measure than they are to the format of the questions. Simply stated, there is greater evidence that it is mathematics and English that are being measured rather than the ability to answer multiple choice or essay or rubric-scored formats.In particular, the evidence suggests that in the ranges of skills needed to pass the Regents (commencement-level) examination or to achieve competency (proficiency level 3) performance on the fourth and eighth grade tests, the skill intended to be measured is the predominant skill being measured.Two appendixes contain details of the analyses. (Contains 15 tables and 6 references.) (SLD)