Rebecca Oh

Numerous studies pertaining to sensitive survey questions have proposed the notion that responses to sensitive questions depend on the context and circumstances in which individuals respond. In particular, many pieces of literature have suggested lower levels of social desirability biases in self-administered electronic surveys than in other modes. Since the transition from paper and telephone modes of data collection to electronic modes has been become increasingly prevalent, this study focused on measurement differences of sensitive survey items – specifically mental health – of different computer devices (smartphone, tablet, or PC). Notably, we examined response rates and determined whether there is a statistically significant difference between devices. We further explored measurement differences by performing a chi-square test for goodness-of-fit to find if device and mental health severity classifications are independent. We found that there is no statistically significant difference in response rates between smartphones, tablets, and PCs, and the proportion of respondents who disclose their mental health severity are not distributed differently across various computer devices.