|
Field
Abstract
|
Before
This study is composed of two parts.
In part 1 we study the effect of incentives and payment mechanisms in preference elicitation. We vary incentives at four levels (0.2% chance of subject getting paid vs. 1% chance of subject getting paid vs. 50% chance of subject getting paid vs. 100% chance of subject getting paid) and payment mechanism at five levels (pay one randomly/POR vs. pay all correlated with one draw at the end for all decisions/PAC vs. PAC divided by number of decisions vs. pay all at the end with independent draws for each decision/PAI vs. PAI divided by number of decisions). Incentives and payment mechanisms are varied between subjects. We use an induced value BDM mechanism to elicit preferences that varies within subjects the Induced Value ($1 and $3) with which subjects are endowed and the support of the price distribution ($4, $5 and $6) from which random prices are drawn. Subjects receive a $2.5 show up fee plus their earnings from part 1.
In part 2 of this study we elicit preferences for a conventional and a Criollo steak. We endow subjects with the conventional steak and ask them to bid to exchange the conventional steak with a Criollo steak. We elicit preferences for the Criollo steak with the BDM mechanism and vary between subjects the support of the distribution from which random prices are drawn ($4 and $5). Subjects are assigned to one of the following treatments: a) hypothetical treatment b) real treatment c) a treatment where subjects are told they have a 10% chance of actually having their decision realized d) a treatment where subjects are told that 50 out of 500 subjects will have their decision realized e) a treatment where subjects are told they have a 1% chance of having their decision realized. For subjects that have their decisions realized, we will ship steaks to their address via 2-day ground coolers dry ice via UPS/Fedex. Only those subjects that have their decisions realized, are endowed with $5.
For a subsample of subjects (500), we also test whether submitting a bid with a slider or a box makes a difference. We added two tasks similar to Part 1. Subjects bid in an induced value BDM ($2) with a support that varies at two levels ($3, $4). One of the two tasks is randomly drawn for payment. Subjects are randomized either to a treatment with sliders or a treatment where they use a box to submit their bid.
|
After
This study is composed of two parts.
In part 1 we study the effect of incentives and payment mechanisms in preference elicitation. We vary incentives at five levels (hypothetical vs. 0.2% chance of subject getting paid vs. 1% chance of subject getting paid vs. 50% chance of subject getting paid vs. 100% chance of subject getting paid) and payment mechanism at five levels (pay one randomly/POR vs. pay all correlated with one draw at the end for all decisions/PAC vs. PAC divided by number of decisions vs. pay all at the end with independent draws for each decision/PAI vs. PAI divided by number of decisions). Incentives and payment mechanisms are varied between subjects. We use an induced value BDM mechanism to elicit preferences that varies within subjects the Induced Value ($1 and $3) with which subjects are endowed and the support of the price distribution ($4, $5 and $6) from which random prices are drawn. Subjects receive a $2.5 show up fee plus their earnings from part 1.
In part 2 of this study we elicit preferences for a conventional and a Criollo steak. We endow subjects with the conventional steak and ask them to bid to exchange the conventional steak with a Criollo steak. We elicit preferences for the Criollo steak with the BDM mechanism and vary between subjects the support of the distribution from which random prices are drawn ($4 and $5). Subjects are assigned to one of the following treatments: a) hypothetical treatment b) real treatment c) a treatment where subjects are told they have a 10% chance of actually having their decision realized d) a treatment where subjects are told that 50 out of 500 subjects will have their decision realized e) a treatment where subjects are told they have a 1% chance of having their decision realized. For subjects that have their decisions realized, we will ship steaks to their address via 2-day ground coolers dry ice via UPS/Fedex. Only those subjects that have their decisions realized, are endowed with $5.
For a subsample of subjects (500), we also test whether submitting a bid with a slider or a box makes a difference. We added two tasks similar to Part 1. Subjects bid in an induced value BDM ($2) with a support that varies at two levels ($3, $4). One of the two tasks is randomly drawn for payment. Subjects are randomized either to a treatment with sliders or a treatment where they use a box to submit their bid.
|
|
Field
Trial End Date
|
Before
September 10, 2022
|
After
September 20, 2022
|
|
Field
Last Published
|
Before
August 25, 2022 04:57 PM
|
After
September 13, 2022 01:23 PM
|
|
Field
Intervention (Public)
|
Before
This study is composed of two parts.
In part 1 subjects are randomized on a between-subjects basis in one of the treatments of a 3x5 design. We vary incentives at three levels (1% chance of subject getting paid vs. 50% chance of subject getting paid vs. 100% chance of subject getting paid) and payment mechanism at five levels (pay one randomly/POR vs. pay all correlated with one draw at the end for all decisions/PAC vs. PAC divided by number of decisions vs. pay all at the end with independent draws for each decision/PAI vs. PAI divided by number of decisions). We use an induced value BDM mechanism to elicit preferences that varies within subjects the Induced Value ($1 and $3) with which subjects are endowed and the support of the price distribution ($4, $5 and $6) from which random prices are drawn. Subjects receive a $2.5 show up fee plus their earnings from part 1.
In part 2 of this study we elicit preferences for a conventional and a Criollo steak. We endow subjects with 5$ and the conventional steak and ask them to bid to exchange the conventional steak with a Criollo steak. We elicit preferences for the Criollo steak with the BDM mechanism and vary between subjects the support of the distribution from which random prices are drawn ($4 and $5). Subjects are assigned to one of the following treatments: a) hypothetical treatment b) real treatment c) a treatment where subjects are told they have a 10% chance of actually having their decision realized d) a treatment where subjects are told that 50 out of 500 subjects will have their decision realized. For subjects that have their decisions realized, we will ship steaks to their address via 2-day ground coolers dry ice via UPS/Fedex. Only those subjects that have their decisions realized, are endowed with $5.
After data collection for the above design was completed, we added a treatment for Part 1 for incentives and subjects were given a 0.2% (1 in 500 chance) of having their decisions realized and payment mechanism was varied between subjects at five levels (pay one randomly/POR vs. pay all correlated with one draw at the end for all decisions/PAC vs. PAC divided by number of decisions vs. pay all at the end with independent draws for each decision/PAI vs. PAI divided by number of decisions).
We also added one treatment in Part 2 where subjects had a 1% chance of having their decisions being realized.
For this additional treatments, we added a Part 3 in order to test whether submitting a bid with a slider or a box makes a difference. To this regard, we added two tasks similar to Part 1. Subjects bid in an induced value BDM ($2) with a support that varies at two levels ($3, $4). One of the two tasks is randomly drawn for payment. Subjects are randomized either to a treatment with sliders or a treatment where they use a box to submit their bid.
|
After
This study is composed of two parts.
In part 1 subjects are randomized on a between-subjects basis in one of the treatments of a 3x5 design. We vary incentives at three levels (1% chance of subject getting paid vs. 50% chance of subject getting paid vs. 100% chance of subject getting paid) and payment mechanism at five levels (pay one randomly/POR vs. pay all correlated with one draw at the end for all decisions/PAC vs. PAC divided by number of decisions vs. pay all at the end with independent draws for each decision/PAI vs. PAI divided by number of decisions). We use an induced value BDM mechanism to elicit preferences that varies within subjects the Induced Value ($1 and $3) with which subjects are endowed and the support of the price distribution ($4, $5 and $6) from which random prices are drawn. Subjects receive a $2.5 show up fee plus their earnings from part 1.
In part 2 of this study we elicit preferences for a conventional and a Criollo steak. We endow subjects with 5$ and the conventional steak and ask them to bid to exchange the conventional steak with a Criollo steak. We elicit preferences for the Criollo steak with the BDM mechanism and vary between subjects the support of the distribution from which random prices are drawn ($4 and $5). Subjects are assigned to one of the following treatments: a) hypothetical treatment b) real treatment c) a treatment where subjects are told they have a 10% chance of actually having their decision realized d) a treatment where subjects are told that 50 out of 500 subjects will have their decision realized. For subjects that have their decisions realized, we will ship steaks to their address via 2-day ground coolers dry ice via UPS/Fedex. Only those subjects that have their decisions realized, are endowed with $5.
After data collection for the above design was completed, we added a treatment for Part 1 for incentives and subjects were given a 0.2% (1 in 500 chance) of having their decisions realized and payment mechanism was varied between subjects at five levels (pay one randomly/POR vs. pay all correlated with one draw at the end for all decisions/PAC vs. PAC divided by number of decisions vs. pay all at the end with independent draws for each decision/PAI vs. PAI divided by number of decisions).
We also added one treatment in Part 2 where subjects had a 1% chance of having their decisions being realized.
For this additional treatments, we added a Part 3 in order to test whether submitting a bid with a slider or a box makes a difference. To this regard, we added two tasks similar to Part 1. Subjects bid in an induced value BDM ($2) with a support that varies at two levels ($3, $4). One of the two tasks is randomly drawn for payment. Subjects are randomized either to a treatment with sliders or a treatment where they use a box to submit their bid.
After collecting data for these additional treatments we added a hypothetical treatment for Part 1.
|
|
Field
Intervention End Date
|
Before
September 10, 2022
|
After
September 20, 2022
|
|
Field
Experimental Design (Public)
|
Before
This study is composed of two parts.
In part 1 subjects are randomized on a between-subjects basis in one of the treatments of a 3x5 design. We vary incentives at three levels (1% of subjects gets paid vs. 50% of subjects vs. 100% of subjects) and payment mechanism at five levels (pay one randomly/POR vs. pay all correlated with one draw at the end for all decisions/PAC vs. PAC divided by number of decisions vs. pay all at the end with independent draws for each decision/PAI vs. PAI divided by number of decisions). We use an induced value BDM mechanism to elicit preferences that varies within subjects the Induced Value ($1 and $3) with which subjects are endowed and the support of the price distribution ($4, $5 and $6) from which random prices are drawn. Subjects receive a $2.5 show up fee plus their earnings from part 1.
In part 2 of this study we elicit preferences for a conventional and a Criollo steak. We endow subjects with 5$ and the conventional steak and ask them to bid to exchange the conventional steak with a Criollo steak. We elicit preferences for the Criollo steak with the BDM mechanism and vary between subjects the support of the distribution from which random prices are drawn ($4 and $5). Subjects are assigned to one of the following treatments: a) hypothetical treatment b) real treatment c) a treatment where subjects are told they have a 10% chance of actually having their decision realized d) a treatment where subjects are told that 50 out of 500 subjects will have their decision realized. For subjects that have their decisions realized, we will ship steaks to their address via 2-day ground coolers dry ice via UPS/Fedex. Only those subjects that have their decisions realized, are endowed with $5.
After data collection for the above design was completed, we added a treatment for Part 1 for incentives and subjects were given a 0.2% (1 in 500 chance) of having their decisions realized and payment mechanism was varied between subjects at five levels (pay one randomly/POR vs. pay all correlated with one draw at the end for all decisions/PAC vs. PAC divided by number of decisions vs. pay all at the end with independent draws for each decision/PAI vs. PAI divided by number of decisions).
We also added one treatment in Part 2 where subjects had a 1% chance of having their decisions being realized.
For this additional treatments, we added a Part 3 in order to test whether submitting a bid with a slider or a box makes a difference. To this regard, we added two tasks similar to Part 1. Subjects bid in an induced value BDM ($2) with a support that varies at two levels ($3, $4). One of the two tasks is randomly drawn for payment. Subjects are randomized either to a treatment with sliders or a treatment where they use a box to submit their bid.
|
After
This study is composed of two parts.
In part 1 subjects are randomized on a between-subjects basis in one of the treatments of a 3x5 design. We vary incentives at three levels (1% of subjects gets paid vs. 50% of subjects vs. 100% of subjects) and payment mechanism at five levels (pay one randomly/POR vs. pay all correlated with one draw at the end for all decisions/PAC vs. PAC divided by number of decisions vs. pay all at the end with independent draws for each decision/PAI vs. PAI divided by number of decisions). We use an induced value BDM mechanism to elicit preferences that varies within subjects the Induced Value ($1 and $3) with which subjects are endowed and the support of the price distribution ($4, $5 and $6) from which random prices are drawn. Subjects receive a $2.5 show up fee plus their earnings from part 1.
In part 2 of this study we elicit preferences for a conventional and a Criollo steak. We endow subjects with 5$ and the conventional steak and ask them to bid to exchange the conventional steak with a Criollo steak. We elicit preferences for the Criollo steak with the BDM mechanism and vary between subjects the support of the distribution from which random prices are drawn ($4 and $5). Subjects are assigned to one of the following treatments: a) hypothetical treatment b) real treatment c) a treatment where subjects are told they have a 10% chance of actually having their decision realized d) a treatment where subjects are told that 50 out of 500 subjects will have their decision realized. For subjects that have their decisions realized, we will ship steaks to their address via 2-day ground coolers dry ice via UPS/Fedex. Only those subjects that have their decisions realized, are endowed with $5.
After data collection for the above design was completed, we added a treatment for Part 1 for incentives and subjects were given a 0.2% (1 in 500 chance) of having their decisions realized and payment mechanism was varied between subjects at five levels (pay one randomly/POR vs. pay all correlated with one draw at the end for all decisions/PAC vs. PAC divided by number of decisions vs. pay all at the end with independent draws for each decision/PAI vs. PAI divided by number of decisions).
We also added one treatment in Part 2 where subjects had a 1% chance of having their decisions being realized.
For this additional treatments, we added a Part 3 in order to test whether submitting a bid with a slider or a box makes a difference. To this regard, we added two tasks similar to Part 1. Subjects bid in an induced value BDM ($2) with a support that varies at two levels ($3, $4). One of the two tasks is randomly drawn for payment. Subjects are randomized either to a treatment with sliders or a treatment where they use a box to submit their bid.
After collecting data for these additional treatments we added a hypothetical treatment for Part 1.
|
|
Field
Planned Number of Clusters
|
Before
1500 subjects + 500 subjects for the additional treatments after collecting data for the 1500 subjects
|
After
1500 subjects + 500 subjects for the additional treatments after collecting data for the 1500 subjects + 500 subjects for the hypothetical treatment
|
|
Field
Planned Number of Observations
|
Before
Part 1: we will recruit 1500 subjects that will be evenly spread along the 15 cells of our experimental design (100 subjects per treatment). Each subject will make 6 decisions. The additional 500 subjects are split between 5 cells of our experimental design that varies the payment mechanism.
Part 2: Subjects from Part 1, will make an additional decision by bidding for having a conventional steak exchanged to a Criollo steak.
Part 3: The additional 500 subjects.
|
After
Part 1: we will recruit 1500 subjects that will be evenly spread along the 15 cells of our experimental design (100 subjects per treatment). Each subject will make 6 decisions. The additional 1000 subjects (for the hypothetical and the 0.2% chance treatment) are split between 5 cells of our experimental design that varies the payment mechanism.
Part 2: Subjects from Part 1, will make an additional decision by bidding for having a conventional steak exchanged to a Criollo steak.
Part 3: The additional 500 subjects.
|