Incentives for subjects in internet experiments

Internet experiments are a new and convenient way for reaching a large subject pool. Yet, providing incentives to subjects can be a tricky design issue. One cost effective and simple method is the publication of a high score (as in computer games). We test whether a high score provides adequate and non-distortionary incentives by comparing it to the usual performance based incentives. We find significant differences and conclude that high scores are not always appropriate as an incentive device. Performance based financial incentives seem to be required also in internet experiments.