Experimental Design
First, we provide respondents two lotteries, among which one showing the probability of winning and one showing no probability of winning.
Second, we randomly assign the second type of lottery to different groups of participants. More precisely, the first type of lottery, which is refered to number lottery, is drawing a ticket numbered from 1-100 from a box with 100 tickets. On each number lottery, there is a set of numbers, which varies from lottery to lottery, e.g. Lottery: winning numbers are 1, ..., 20.
For the second type of lottery, we follow Li (2007) and provide three phrases in Chinese (Beijing dialect, Shandong dialect and Shanghai dialect, respectively). We randomly assign these phrase lotteries to three groups. On each phrase lottery, there is a set of winning meansings which varies from lottery to lottery. We can then obtain their preferences about uncertainty and risk by using the turning point of choice between two lotteries.
Third, we ask respondents their consumption-saving behaviors.