Assessing the Development of Math and Science Attitudes: A RI-CLPM Study

Abstract

Attitudes towards math and science are important contributors to later science, technology, engineering, and math (STEM) participation. As such, policymakers planning to target these attitudes may benefit from understanding how math and science attitudes develop and how they might influence each other across development. Using data from the Longitudinal Study of American Youth (LSAY), this thesis examines the development of math and science attitudes from grade 7-12 in a sample of approximately 3,000 students. Using the random intercept cross-lagged panel model (RI-CLPM), results show that math and science attitudes develop in isolation. Early math attitudes tend to predict later math attitudes, and earlier science attitudes predict later science attitudes, with little cross over. Results further indicate that male and female students show a different pattern of development for math and science attitudes. Female students show stronger auto-regressive paths for both math and science (particularly after grade nine), while male students show a stronger between-subject correlation of math and science attitudes. Generally, these results demonstrate that the development of math and science attitudes is isolated, and that earlier within-subject attitudes are important in shaping later attitudes of the same domain. With respect to gender, policymakers should consider targeted approaches for cultivating mathand science attitudes given the different developmental trajectories.