|
Field
Abstract
|
Before
We design a survey experiment 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. 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 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?
|
|
Field
Trial Start Date
|
Before
December 25, 2021
|
After
December 31, 2021
|
|
Field
Last Published
|
Before
December 24, 2021 05:05 PM
|
After
December 24, 2021 08:29 PM
|
|
Field
Intervention (Public)
|
Before
We measure economic rationality of individual preferences. As such, we do not have an experimental intervention. Rather, we rely in natural variation on subject's characteristics.
|
After
We measure economic rationality of individual preferences. As such, we do not have an experimental intervention. Rather, we rely in natural variation on subject's characteristics.
|
|
Field
Intervention Start Date
|
Before
December 25, 2021
|
After
December 31, 2021
|
|
Field
Primary Outcomes (End Points)
|
Before
The key outcome variables in this experiment are consistency scores measured by the choice data, as indicators of economic rationality in risk preference, social preference and food preference.
|
After
We use revealed preferences analysis to check whether subjects’ behavior comply with utility maximization hypothesis. It will be done by checking whether individual behavior is consistent with Generalized Axiom of Revealed Preferences (GARP) and measuring the extent to which it violates GARP. We use various indices of measuring GARP violations, including Afriat’s Critical Cost Efficiency Index (CCEI) and Money Pump Index (MPI). This will be separately done in domains of decision-making under risk preference, social preference and food preference.
|
|
Field
Primary Outcomes (Explanation)
|
Before
1. Rationality in food preference
Subjects need to allocate a fixed expenditure of 50 RMB between vegetables and meat to choose their best buying amount in 22 rounds. With these decisions, we can observe violations of GARP (Generalized Axiom of Revealed Preferences) and compute their Critical Cost Effective Index (CCEI).
2. Ratioanlity in risk preference
Following Choi et al. (2007, 2014), Subjects have 100 points to allocate between two accounts where a point has different RMB value in 22 rounds. With these decisions, we can observe violations of GARP (Generalized Axiom of Revealed Preferences) and compute their Critical Cost Effective Index (CCEI).
3. Ratioanlity in social preference
Following Andreoni&Miller (2002), Subjects have 100 points to allocate between themselves and another subject where a point has different RMB value in 22 rounds. With these decisions, we can observe violations of GARP (Generalized Axiom of Revealed Preferences) and compute their Critical Cost Effective Index (CCEI).
|
After
1. Rationality in food preference
Subjects need to allocate a fixed expenditure of 50 RMB between vegetables and meat to choose their best buying amount in 22 rounds. With these decisions, we can observe violations of GARP (Generalized Axiom of Revealed Preferences) and compute their Critical Cost Effective Index and Money Pump Index.
2. Ratioanlity in risk preference
Following Choi et al. (2007, 2014), Subjects have 100 points to allocate between two accounts where a point has different RMB value in 22 rounds. With these decisions, we can observe violations of GARP (Generalized Axiom of Revealed Preferences) and compute their Critical Cost Effective Index and Money Pump Index.
3. Ratioanlity in social preference
Following Andreoni&Miller (2002), Subjects have 100 points to allocate between themselves and another subject where a point has different RMB value in 22 rounds. With these decisions, we can observe violations of GARP (Generalized Axiom of Revealed Preferences) and compute their Critical Cost Effective Index and Money Pump Index.
|
|
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. 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.
|
|
Field
Sample size (or number of clusters) by treatment arms
|
Before
300 consumers
|
After
We don't use intervention treatment.
|
|
Field
Keyword(s)
|
Before
Behavior, Welfare
|
After
Behavior
|