Internet-based attitude assessment: does gender affect measurement equivalence?

As researchers and practitioners increasingly turn to the Internet as a data collection medium, investigating the differential effects of administration mode on data quality becomes increasingly important. To date, no research has investigated whether data gathered from differing subgroups demonstrate measurement equivalence across Internet and paper-and-pencil administration modes despite the large literature suggesting that males and females differ in terms of computer anxiety. The present study, employing a repeated measures design, investigated systematic cross-mode differences in responding as a function of gender. Results demonstrate that both sexes use the same psychological metric when providing satisfaction ratings regardless of Internet or paper-and-pencil administration (equivalent factor structure and factor loadings). Furthermore, there were no statistically significant differences in scale/item means over the modes of administration according to gender. These results suggest that gender does not lead to a negative impact on cross-mode measurement equivalence for self-report ratings.