Differential gender performance on math assessment has received continuing attention in the United States. However, research on math gender differences has been thwarted by the lack of rigorous methodology and the lack of generalizable findings. This book, investigates the math gender issue among fifteen-year olds in the U.S, using multi-year data from the Programme for International Student Assessment (PISA), one of the largest international comparison studies. The analyses take advantage of advanced measurement models such as item response models, latent regression models, and mixture models. The results shed light on how various cognitive and noncognitive (i.e., interests, motivation, learning strategies) factors underlie the differential math performance between males and females. This book should be helpful to researchers interested in math gender research and also to researchers interested in the application of advanced measurement models in educational and psychological research.