Family educational pressure, mathematics anxiety, and psychosomatic symptoms: evidence from selected Chinese jurisdictions

Source abstract

Adolescent psychological well-being is increasingly recognized as an important dimension of educational quality, particularly in achievement-oriented educational systems where family expectations may be experienced as pressure rather than support. This study examines the relationships among perceived family educational pressure, mathematics anxiety, and psychosomatic symptoms using data from PISA 2022. The initial data extraction covered the complete PISA 2022 public-use student records for comparable Confucian-heritage East Asian systems including Hong Kong (China), Macao (China), and Chinese Taipei ( N = 16,148). Because psychosomatic-symptom data were unavailable for Chinese Taipei and complete information was required across the focal variables, covariates, and survey weights, the final mediation sample comprised 4,994 students from Hong Kong and Macao. A design-based mediation analysis was conducted using the PISA final student weight and all 80 BRR-Fay replicate weights, controlling for socioeconomic status, gender, grade level, and immigrant background. Family educational pressure was positively associated with mathematics anxiety (B = 0.187, p < 0.001), and mathematics anxiety was positively associated with psychosomatic symptoms after adjustment for family pressure and covariates (B = 0.159, p < 0.001). Family pressure also retained a positive direct association with psychosomatic symptoms (B = 0.122, p < 0.001). The design-based indirect association through mathematics anxiety was statistically distinguishable from zero (ab = 0.0303, 95% CI [0.0217, 0.0388]). The pattern of findings was positive across both participating economies and remained evident after additional adjustment for mathematics achievement using PISA plausible values. Overall, the findings indicate a modest cross-sectional indirect association linking perceived family educational pressure with psychosomatic symptoms through mathematics anxiety. Given the cross-sectional design, the results are interpreted as associations rather than causal effects.