This study aims to test the hybrid model structured by the variables of instrumental motivation for mathematics, mathematics self-concept, mathematics self-efficacy, mathematics anxiety and mathematics interest that are considered to have affected mathematics behavior and mathematics literacy of the PISA 2012 Turkey sample, and to examine if the model varies in terms of gender and school type (general high school, Anatolian school and vocational high school). The population of this correlational study consists of 4848 students who attended PISA 2012 from Turkey and the sample consists of 15-year-old 1441 students who took B form of the Student Questionnaire. PISA 2012 "Student Questionnaire-B Form" and "Mathematics Literacy Test" were used as data collecting tools in the research. When fit indices were analyzed, it was found that model-data fit was achieved for the hybrid model structured at the first phase and the model didn't vary according to gender and school type. In addition, when path coefficients were examined, it was seen that mathematics self-efficacy was the best predictor of the mathematics literacy variable and mathematics interest was the best predictor of the mathematics behavior variable. It was observed that mathematics self-efficacy had a great positive effect in predicting mathematics literacy, and mathematics interest had a great positive effect in predicting mathematics behavior. The results of the study revealed that structured model explained 44% of the mathematics behavior and 39% of the mathematics literacy.