A comparative study of lotteries-evaluation in class and on the Web

We report the results of an experiment in which 135 students were asked to bid buying prices for five simple lotteries. 65 subjects were asked to complete the evaluation forms in class; the other 70 subjects were asked to complete the questionnaire within 24 hours on the Web. The subjects for both groups were carefully selected to avoid a possible selection-bias. The main results are: (1) bids on the Web are significantly higher than bids in class; (2) the standard deviations of the bids are significantly higher on the Internet. We also find some differences across gender in the response to the electronic medium. Our results suggest that subjects "risk aversion" might be lower on the Web and that the Internet medium might increase the noise in experimental data.