When Incentives Matter Too Much: Explaining Significant Responses to Irrelevant Information