|
Field
Abstract
|
Before
We conduct a survey experiment in consumers of supermarkets in China to measure people’s economic rationality in risk preference domain, social preference domain and food preference domain and compare them to check consistency between these three rationalities. Our motivation stems from the result of our previous survey experiment, which finds that the inconsistency between rationality in risk preference in the lab and consuming behavior in the field from scanner data. Therefore, we design another survey experiment in supermarket consumers to check the underlying mechanism of this inconsistency and check the cross validity of rationality. Firstly, we design a food task to simulate a supermarket environment where subjects optimize between meat and vegetables under a fixed expenditure of 50 yuan, from which we can measure rationality in the food preference that is similar to consuming behavior in the field. Secondly, we design risk task and social task to measure subjects' rationality in risk preference and social preference separately, from which we can measure rationality in the lab. Thirdly, we will investigate the consistency between rationality in the food preference and risk/social preference: does the inconsistency between lab and field stems from the external invalidity of rationality in the lab or just problems in the scanner data? Then we will investigate the consistency between rationality in the risk preference and social preference: Is there any cross validity between risk preference and social preference in the lab?
|
After
We conduct a survey experiment for supermarkets consumers in China to measure individual's economic rationality in risk preference, social preference domain and food preference domains. Our prior research has shown the inconsistency between rationality in risk preference measured in the survey experiment and rationality stemmed from scanner data. Therefore, we design this survey experiment to examine the underlying mechanism of such inconsistency and validate the cross validity of rationality measurements.
|
|
Field
Experimental Design (Public)
|
Before
We design a survey experiment in supermarkets which will take approximately 25 minutes. The experiment consists of four sections. In Section 1, each participant has to make choices in the three tasks. In the food task, subjects are asked to allocate a budget between meat and vegetables in 22 questions. In the risk task, subjects are asked to allocate a budget between two risky assets, each one of which obtains with a probability of 0.5. The assets have different payoffs, so that allocations give an indication of risk preferences. There are 22 questions in this task. In the social task, subjects are asked to allocate a budget between themselves and another subject. The both have different payoffs, so that allocations give an indication of social preferences. The order of these 3 task are random, and computer randomly chooses one round from the 66 rounds of three tasks and participant will get rewarded depending on the outcome of this chosen round.
In Section 2, we measure 5 Big personality of subjects. In Section 3, we test subjects' IQ by raven Test. Lastly, we ask participants for their demographic and economic information, including gender, education level and other socio-economic and sociodemographic characteristics.
|
After
We design a survey experiment in supermarkets which will take approximately 25 minutes. First, each participant makes choices in three tasks. In the food task, subjects are asked to allocate a budget between meat and vegetables in 22 questions. In the risk task, subjects are asked to allocate a budget between two risky assets.. The assets have different payoffs, so that allocations give an indication of risk preferences. There are 22 questions in this task as well. In the social preference task, subjects are asked to allocate a budget between themselves and another subject. Again, there are 22 questions for the social preference tasks.. We randomize the order of these three tasks and computer randomly chooses one round from the 66 rounds of three tasks for payment. Additionally, we also measure individuals’ big five personality traits, IQ and collect their demographic information in the post-experiment survey.
|