The Internal Structure of New Attitudes towards Statistics Inventory among Postgraduate Students in Medical and Health Sciences

Universiti Sains Malaysia Attitudes toward Statistics (USM-AS) is a newly developed, selfadministered inventory for measuring attitudes toward statistics among postgraduate students. The USM-AS consists of three factors (affect, cognition, and behaviour) and 58 items. This study was aimed to validate the USM-AS and provide evidence of its construct validity by investigating its internal structure. This study consists of four validation studies, which were conducted consecutively among postgraduate students (Master’s degree and PhD students) in the medical and health sciences. The revised USM-AS, consisting of three factors and 12 items, was found to have good construct validity, which was demonstrated by good model fit, high factor loadings (0.62 to 0.92), high construct reliability (0.84 to 0.89), good discrimination between factors and good test-retest reliability (intraclass correlation coefficients = 0.63 to 0.82). However, validity was limited to postgraduate students in the medical and health sciences in Malaysia. Further validation studies among postgraduate students in other scientific fields are recommended to provide additional construct validity evidence of the USMAS.