摘要：The interdisciplinary curriculum of science, technology, engineering, math, and project-based learning (STEM-PBL) has currently received deep contemplation for its diverse benefits. This major attention has been focused on optimizing the curriculum by developing tools for assessing the diverse aspects of STEM in relation to teaching methods (e.g., PBL, teacher-centered learning, and game-based teaching), and students’ attitude. Following the principle of fallibility, the current research examined psychometric properties of the scale of students’ attitudes toward STEM-PBL. Also, the power of predictability of academic achievement by means of students’ attitude, and the mediating role of demographic factors were aimed to explore. The scale was completed by 337 students who were selected randomly. Using confirmatory factor analysis, Graded Response Modeling (GRM) based on Item Response Theory (IRT) provided technical informative details about each item of scale. Examining the items of scale displays reasonable psychometric properties including the validity of Student Attitude toward STEM-PBL Questionnaire. However, the analysis revealed that the removal of one of the items (item 22) may result in a better fit of the questionnaire. The results also showed that attitude toward STEM-PBL can predict students’ academic achievement using grade point average (GPA). Exploring the role of some mediating and demographic factors in this prediction revealed that self-learning had a high significant unique contribution in predicting GPA among respondents. The results displayed the significant association between GPA and students’ father education level. Also, students whose GPA and their fathers’ educational levels were high showed more positive attitude toward self-learning. Furthermore, the time of using technology and hands-on activity as a component of the students’ attitude toward STEM-PBL showed indirect relationship.
