Objectives: This study aims to analyze various psychological propensities with fuzzy questionnaires in physical education. Motivations: Fuzzy questionnaires with fuzzy statistical analysis can better reflect people's psychological thinking. Method: Fuzzy descriptive statistics, fuzzy correlation coefficients and fuzzy regression models. Finding: Comparing among three scales, fuzzy Kaiser-Meyer-Olkin (KMO), Cronbach's α and traditional one: fuzzy Mastery- Avoidance had a positive correlation with self-esteem. Fuzzy positive self-esteem significantly impacted predictive Mastery-Approach (AF_MA) and Mastery-Avoidance (AF_M_a); while traditional positive self-esteem significantly impacted predictive Performance-Avoidance (A_P_a) and Mastery-Avoidance (AF_M_a). Further, fuzzy and traditional negative self-esteem significantly impacted predictive AF_MA, simultaneously fuzzy negative self-esteem significantly impacted predictive AF_MA. Only fuzzy negative self-esteem significantly impacted predictive AF_MA. Innovations: Traditional psychological examination technology has been used for many years, while a new approach with fuzzy statistical analysis on artificial intelligence technology will promote the quality of sampling survey and marketing information. Value: This study offered new examination skill in analyzing psychological propensity in physical education.