Information Avoidance and Ingroup Favoritism: The Case of (Invisible) Vaccination Status

The allocation game consists of two main conditions (Self/Other and Other/Other), and within each condition, there are three subconditions (Ingroup, Outgroup, and Uncertain).

1. Self/Other Condition
• The participant is always selected as the decision-maker.
• Player 1 is the participant (You), and Player 2 is Z.
• Player 1 (You) is always vaccinated, meaning you have received at least three doses of the COVID-19 vaccine and intend to receive vaccinations in future pandemics.

1.1. Ingroup Condition
• Player 2 (Z) is also vaccinated, meaning they have received at least three doses of the COVID-19 vaccine and intend to receive vaccinations in future pandemics.

1.2. Outgroup Condition
• Player 2 (Z) is unvaccinated, meaning they have never received any dose of the COVID-19 vaccine and do not intend to receive vaccinations in future pandemics.

1.3. Uncertain Condition
• Player 2 (Z) has a 50% probability of being vaccinated and a 50% probability of being unvaccinated.
• Before making the allocation decision, the participant is offered a cost-free opportunity to verify Player 2’s vaccination status.
• The participant may choose to utilize this opportunity or to ignore it.

2. Other/Other Condition
• The participant is always selected as the decision-maker.
• Player 1 is Y, and Player 2 is Z.
• Player 1 (Y) is always vaccinated.

2.1. Ingroup Condition
• Player 2 (Z) is also vaccinated.

2.2. Outgroup Condition
• Player 2 (Z) is unvaccinated.

2.3. Uncertain Condition
• Player 2 (Z) has a 50% probability of being vaccinated and a 50% probability of being unvaccinated.
• Before making the allocation decision, the participant is offered a cost-free opportunity to verify Player 2’s vaccination status.
• The participant may choose to utilize this opportunity or to ignore it.

2025-03-14

2025-03-21

1. A binary variable indicating whether the participant selected the selfish option in the allocation game.

2. A binary variable indicating whether the participant avoided the opportunity to verify Player 2’s vaccination status.

1. Allocation Game:
o In the allocation game, experimental participants act as decision-makers and determine how to allocate 100 JPY between Player 1 and Player 2.
o As described in the experimental design, in the Self/Other condition, Player 1 is the participant themselves, and Player 2 is another individual. In contrast, in the Other/Other condition, both Player 1 and Player 2 are other individuals.
o The game presents two allocation options as follows, with Option B defined as the selfish choice in this study:
 Option A: Player 1 – 50 JPY, Player 2 – 50 JPY
 Option B: Player 1 – 70 JPY, Player 2 – 30 JPY

2. Opportunity to Verify Player 2’s Vaccination Status:
o As described in the experimental design, in the Uncertain condition, Player 2’s vaccination status is not known to the participant before making the allocation decision.
o However, participants are offered a cost-free opportunity to verify Player 2’s vaccination status before making their decision. They may choose to utilize this opportunity or to ignore it.
o This study focuses on participants who choose not to verify Player 2’s vaccination status, a behavior referred to as information avoidance.

Regarding Primary Outcome (2), this study further investigates how participants who chose not to verify Player 2’s vaccination status predicted Player 2’s vaccination status.

After the allocation game in the Uncertain condition, participants who chose not to verify Player 2’s vaccination status will be asked the following question:
"Did you predict that Player 2 (Z) was more likely to be vaccinated, or did you predict that they were more likely to be unvaccinated?"

In this study, participants take part in a financially incentivized allocation game. At the beginning of the experiment, the following rules are explained:

• Each participant (You) is grouped with two anonymous individuals, Y and Z, forming a three-person set.
• Y and Z are Japanese survey respondents registered with the same online survey company and reside in different regions across Japan.
• The base participation compensation for Y and Z is the same as that of the participant.
• None of the three individuals know each other, and they will not meet in person during the game.
• One individual is randomly selected as the decision-maker in each round.
• Based on the decision-maker’s choice, two of the three individuals (including the participant, Y, or Z) have the chance to receive an additional monetary reward as bonus points.
• The allocation game is conducted five times.
• After the experiment, one of the five rounds is randomly selected, and the additional compensation is distributed according to the decision made in that round.

The allocation game consists of two main conditions (Self/Other and Other/Other), and within each condition, there are three subconditions (Ingroup, Outgroup, and Uncertain).

1. Self/Other Condition
• The participant is always selected as the decision-maker.
• Player 1 is the participant (You), and Player 2 is Z.
• Player 1 (You) is always vaccinated, meaning you have received at least three doses of the COVID-19 vaccine and intend to receive vaccinations in future pandemics.

1.1. Ingroup Condition
• Player 2 (Z) is also vaccinated, meaning they have received at least three doses of the COVID-19 vaccine and intend to receive vaccinations in future pandemics.

1.2. Outgroup Condition
• Player 2 (Z) is unvaccinated, meaning they have never received any dose of the COVID-19 vaccine and do not intend to receive vaccinations in future pandemics.

1.3. Uncertain Condition
• Player 2 (Z) has a 50% probability of being vaccinated and a 50% probability of being unvaccinated.
• Before making the allocation decision, the participant is offered a cost-free opportunity to verify Player 2’s vaccination status.
• The participant may choose to utilize this opportunity or to ignore it.

2. Other/Other Condition
• The participant is always selected as the decision-maker.
• Player 1 is Y, and Player 2 is Z.
• Player 1 (Y) is always vaccinated.

2.1. Ingroup Condition
• Player 2 (Z) is also vaccinated.

2.2. Outgroup Condition
• Player 2 (Z) is unvaccinated.

2.3. Uncertain Condition
• Player 2 (Z) has a 50% probability of being vaccinated and a 50% probability of being unvaccinated.
• Before making the allocation decision, the participant is offered a cost-free opportunity to verify Player 2’s vaccination status.
• The participant may choose to utilize this opportunity or to ignore it.

Each participant plays five rounds of the allocation game, each under a different condition. As detailed in the experimental design, participants are randomly assigned to one of 20 condition sequences. This approach allows for both between-subject and within-subject analyses. For instance, by comparing differences in the first-round allocation decisions across individuals, it is possible to conduct between-condition comparisons across the six conditions.

Not available

Stratified randomization is conducted based on four strata of gender and age, and participants are randomly assigned to one of the experimental sequences described in the Details section.

Individuals

No

N/A

2,400 vaccinated individuals

Focusing on the condition of the first allocation game out of the five rounds, 400 participants are assigned to each of the following six conditions.

1. Self/Other condition (N=1,200)
1.1. Ingroup condition (n=400)
1.2. Outgroup condition (n=400)
1.3. Uncertain condition (n=400)

2. Other/Other condition (N=1,200)
2.1. Ingroup condition (n=400)
2.2. Outgroup condition (n=400)
2.3. Uncertain condition (n=400)

I used Stata16.1 to conduct a power analysis. Our goal was to obtain 0.80 power to detect a delta of 0.1001 with a standard deviation of 0.4845 at 0.050 alpha error probability. The delta and standard deviation were based on the result of our pilot experiment. More specifically, I used the lower bound of one standard deviation of the estimated value as the delta. The estimated sample size for each group is 370, and thus I set the sample size for each group to 400.