This proposed research uses a multidimensional approach to validating the constructs of student achievement in high school mathematics and science. It elaborates and indicator of student achievement to enhance the validity of large scale education surveys. Specifically it identifies knowledge and reasoning components of mathematics and reasoning test that are both multiple choice and constructed response using the NELS:88 High School Effects Study (HSES). It analyses the constructed response tasks for con students of different gender and ethnicity's: it extends the construct approach to the National Assessment of Education Progress (NAEP) and develops improved procedures for conduction interview studies in item construct validation. It also will develop guidelines for other investigators who apply multidimensional scoring of student test.
