The Consumer Expenditure interview survey consists of a 90 minutes of filter questions about putchases in categories such as, clothing, insurance, vehicle expenses. In most sections, each reported purchase triggers follow-up questions about the amount spent, whom the purchase was for, etc. Respondents may learn that answering yes to these filter questions leads to further questions and begin to misreport and answer no. Such motivated misreorting has been found in other surveys using filter questions. This talk will investigate whether motivated misreporting is taking place in the Consumer Expenditure survey.
Much of my research in the last ten years has explored Motivated Misreporting: the shortcuts respondents take to reduce the length or burden of a survey.
The research for this talk is part of my ASA / NSF / BLS fellowship.