Gender and Cross-Cultural Invariance of Attitudes Toward Statistics Among Non-STEM Undergraduate Students

Because of the wide usage of statistics in almost every field, many universities have incorporated statistics into their undergraduate and graduate curricula. Students taking statistics have often reported negative experiences and an aversion to statistics. Non-STEM students have often reported negative attitudes toward statistics (Chew & Dillon, 2014; Keeley et al., 2008). Researchers have used qualitative and quantitative approaches to measure attitudes toward statistics. The Survey of Attitudes Toward Statistics-36 (SATS-36) scale has been the most widely accepted measure to investigate attitudes toward statistics. Any psychometrically sound measurement instrument requires invariance across groups in its usage. Otherwise, measurement biases could be incorrectly identified as group differences. The three objectives of this study were to (1) validate the six-factor model of the SATS-36 with Sri Lankan undergraduate students, (2) investigate measurement invariance of the factor structure for the proposed model across gender, and (3) investigate the influence of undergraduate students’ cultures on attitudes toward statistics. For objective three the researcher, who is from Sri Lanka, chose to compare the influence of the US culture to the influence of Sri Lankan culture on attitudes toward statistics. Study results revealed that the modified version of the SATS instrument fit well to measure female undergraduate students’ attitudes toward statistics. The modified version of the SATS was partially invariant across gender among undergraduate business students in Sri Lanka. Therefore, the modified model can be used to compare attitudes toward statistics between male and female undergraduate students in the business discipline in Sri Lanka. Male undergraduate students reported more positive attitudes toward statistics for all six factors than their female counterparts did. The difficulty subscale of the instrument did not demonstrate good reliability values for male and female groups. Regarding the cross-cultural comparison of SATS-36 items, Sri Lankan undergraduate students on average reported more positive attitudes toward statistics than the US undergraduate students did. The SATS-36 was noninvariant across Sri Lanka and the US. Specifically, US students’ attitudes towards difficulty of statistics were more negative than Sri Lankan students’ attitudes were. This study also found that amount of work that students expended to learn statistics was not significantly different between both countries’ undergraduate students. These findings are useful to statistics instructors and administrators who aspire to improve their curricula in a way that mitigates negative attitudes toward statistics