|
Field
Abstract
|
Before
This study examines how diversity-based leadership selection impacts decision-making in a team competition. In a university tournament, students participate in an outdoor throwing game where each player makes eight throws, with points ranging from 0 to 20 per throw. Teams consist of two players: Player 1 (P1) and Player 2 (P2), who serves as the team leader. The team score is initially based on three randomly selected throws from each player, while the remaining five throws are kept as replacements. P2, the leader, is given the opportunity to replace the lowest-scoring throw from either P1 or themselves, using one of the replacement throws. However, P2 does not know the actual points of P1’s remaining throws, only their own. P2 makes these decisions for four hypothetical teammates, each varying by gender and social class, with one randomly chosen match being implemented to determine their final team score. Teams in the top 50% of scores participate in a lottery for prizes.
The primary experimental variation lies in whether P2 is informed about the reason for their leadership selection. In the Info treatment, P2 is told that their role as team leader was assigned to meet a diversity goal of 50-50 representation of high and low socioeconomic status leaders. In the No Info treatment, P2 is simply told they are the team leader without mention of the diversity goal. The main outcome measures include: (1) whether P2 chooses to correct P1’s or their own lowest-scoring throw, (2) whose replacement throw they select (P1’s or their own), and (3) how P2’s decisions vary across treatments and depending on P1’s gender and social class. This study aims to shed light on how awareness of diversity goals affects leaders' correction behaviors and decision-making processes in team settings.
|
After
This study examines how diversity-based leadership selection impacts decision-making in a team competition. In a university tournament, students participate in an outdoor throwing game where each player makes eight throws, with points ranging from 0 to 20 per throw. Teams consist of two players: Player 1 (P1) and Player 2 (P2), who serves as the team leader. The team score is initially based on three randomly selected throws from each player, while the remaining five throws are kept as replacements. P2, the leader, is given the opportunity to replace the lowest-scoring throw from either P1 or themselves, using one of the replacement throws. However, P2 does not know the actual points of P1’s remaining throws, only their own. P2 makes these decisions for four hypothetical teammates, each varying by gender and social class, with one randomly chosen match being implemented to determine their final team score. Teams in the top 50% of scores participate in a lottery for prizes.
The primary experimental variation lies in whether P2 is informed about the reason for their leadership selection. In the Info_Class treatment, P2 is told that their role as team leader was assigned to meet a diversity goal of 50-50 representation of high and low socioeconomic status leaders. In the Info_Gender treatment, the diversity goal is on gender (50-50 of men and women). In the No Info treatment, P2 is simply told they are the team leader without mention of the diversity goal. The main outcome measures include: (1) whether P2 chooses to correct P1’s or their own lowest-scoring throw, (2) whose replacement throw they select (P1’s or their own), and (3) how P2’s decisions vary across treatments and depending on P1’s gender and social class. This study aims to shed light on how awareness of diversity goals affects leaders' correction behaviors and decision-making processes in team settings.
|
|
Field
Last Published
|
Before
October 07, 2024 07:15 PM
|
After
October 31, 2024 08:34 AM
|
|
Field
Intervention (Public)
|
Before
The intervention consists of two primary treatments that manipulate the information given to team leaders (P2) about their leadership role. In the **Info treatment**, P2 is informed that they were selected as the team leader to meet a diversity goal set by the tournament organizers, specifically aiming for a 50-50 representation of high and low socioeconomic status (SES) leaders across all teams. This information is intended to make P2 aware of the diversity objective underlying their selection, potentially influencing their decision-making and behavior. In the **No Info treatment**, P2 is simply told that they were selected as the team leader, with no mention of any diversity goal or specific reason for their selection. By comparing behavior between these two treatments, the study aims to explore how the knowledge of being selected for diversity reasons affects leaders’ correction behavior, their choice of whom to correct (themselves or their teammate), and whose replacement throws they choose to use.
|
After
The intervention consists of three primary treatments that manipulate the information given to team leaders (P2) about their leadership role. In the Info_Class treatment, P2 is informed that they were selected as the team leader to meet a diversity goal set by the tournament organizers, specifically aiming for a 50-50 representation of high and low socioeconomic status (SES) leaders across all teams. This information is intended to make P2 aware of the diversity objective underlying their selection, potentially influencing their decision-making and behavior. In the Info_Gender treatment, the information is about the diversity goal focusing on parity in gender (50-50 male and female team leaders). In the No Info treatment, P2 is simply told that they were selected as the team leader, with no mention of any diversity goal or specific reason for their selection. By comparing behavior between these two treatments, the study aims to explore how the knowledge of being selected for diversity reasons affects leaders’ correction behavior, their choice of whom to correct (themselves or their teammate), and whose replacement throws they choose to use.
|
|
Field
Experimental Design (Public)
|
Before
The experiment takes place in the context of a campus-wide tournament at a university, where students register to participate and provide demographic information (gender, social class, parental education). After registration, participants engage in a throwing event in an outdoor game, where they make eight throws that yield points (0, 5, 10, or 20 points per throw). Each participant is randomly paired with another participant after completing the event, with one designated as Player 1 (P1) and the other as Player 2 (P2), who will serve as the team leader.
Each team’s score is determined by three randomly chosen throws from both P1 and P2. The remaining five throws for each player are kept as replacement options. P2, the team leader, is tasked with revising the team’s score by replacing the lowest-scoring throw of either P1 or themselves, using a replacement throw either from their own remaining pile or from P1’s. Crucially, P2 does not know the exact scores of the remaining throws in P1’s pile, introducing an element of uncertainty. P2 is asked which throw they would like to correct and, ultimately, which of the two decisions (replacing their own or P1's throw) they would like to implement. After the revisions, the top 50% of teams based on final scores will be entered into a lottery for team prizes.
The main experimental manipulation involves two treatments: an “Info” treatment, where P2 is informed that their leadership position was assigned due to a diversity goal ensuring 50-50 representation of high and low socioeconomic status (SES) leaders, and a “No Info” treatment, where P2 is simply told they are the leader without being given further details. In addition to this, P2s are asked to make decisions across four potential team matches (strategy method), where P1’s characteristics vary by gender (male or female) and social class (high or low). One of these matches will be randomly chosen to determine the final team.
|
After
The experiment takes place in the context of a campus-wide tournament at a university, where students register to participate and provide demographic information (gender, social class, parental education). After registration, participants engage in a throwing event in an outdoor game, where they make eight throws that yield points (0, 5, 10, or 20 points per throw). Each participant is randomly paired with another participant after completing the event, with one designated as Player 1 (P1) and the other as Player 2 (P2), who will serve as the team leader.
Each team’s score is determined by three randomly chosen throws from both P1 and P2. The remaining five throws for each player are kept as replacement options. P2, the team leader, is tasked with revising the team’s score by replacing the lowest-scoring throw of either P1 or themselves, using a replacement throw either from their own remaining pile or from P1’s. Crucially, P2 does not know the exact scores of the remaining throws in P1’s pile, introducing an element of uncertainty. P2 is asked which throw they would like to correct and, ultimately, which of the two decisions (replacing their own or P1's throw) they would like to implement. After the revisions, the top 50% of teams based on final scores will be entered into a lottery for team prizes.
The main experimental manipulations involves three treatments: an “Info_Class” treatment, where P2 is informed that their leadership position was assigned due to a diversity goal ensuring 50-50 representation of high and low socioeconomic status (SES) leaders, a "Info_Gender" treatment, where the diversity goal is on 50-50 representation of male and female leaders, and a “No Info” treatment, where P2 is simply told they are the leader without being given further details. In addition to this, P2s are asked to make decisions across four potential team matches (strategy method), where P1’s characteristics vary by gender (male or female) and social class (high or low). One of these matches will be randomly chosen to determine the final team.
|
|
Field
Planned Number of Clusters
|
Before
220 participants (at least)
|
After
240 participants (at least)
|
|
Field
Planned Number of Observations
|
Before
220 participants (at least)
|
After
240 participants (at least)
|
|
Field
Sample size (or number of clusters) by treatment arms
|
Before
220 participants (at least)
|
After
240 participants (at least)
|
|
Field
Power calculation: Minimum Detectable Effect Size for Main Outcomes
|
Before
The minimum detectable effect size (MDES) is based on a two-sample comparison with clustering accounted for by the repeated measures (i.e., strategy method of four decisions per participant). We calculated the MDES using a medium effect size, with adjustments for the clustering within participants.
Unit: The main outcome measures are binary decisions regarding whether the leader (Player 2) chooses to replace their own or their teammate's lowest throw. This outcome is measured at the decision level.
Standard Deviation: For binary outcomes (0/1 for whether a correction is made), the standard deviation is assumed to be around 0.5, which is typical when outcomes are evenly distributed across two categories. For continuous outcomes (such as beliefs about remaining throws), we assumed a standard deviation of 1 based on previous studies with similar tasks.
Effect Size (Cohen’s d): Based on the power calculations, the minimum detectable effect size with the adjusted sample of 219 participants, accounting for repeated decisions (clustering within participants) with an intra-class correlation of 0.2, is approximately 0.30. This corresponds to a medium effect size, which is commonly used in social science experiments.
Percentage: This means the study is powered to detect a 30% difference in the probability of correcting a teammate's or their own throws between the Info and No Info treatments, or a difference of similar magnitude in the expected values of replacement throws.
|
After
The minimum detectable effect size (MDES) is based on a two-sample comparison with clustering accounted for by the repeated measures (i.e., strategy method of four decisions per participant). I calculated the MDES using a medium effect size, with adjustments for the clustering within participants.
Unit: The main outcome measures are binary decisions regarding whether the leader (Player 2) chooses to replace their own or their teammate's lowest throw. This outcome is measured at the decision level.
Standard Deviation: For binary outcomes (0/1 for whether a correction is made), the standard deviation is assumed to be around 0.5, which is typical when outcomes are evenly distributed across two categories. For continuous outcomes (such as beliefs about remaining throws), we assumed a standard deviation of 1 based on previous studies with similar tasks.
Effect Size (Cohen’s d): Based on the power calculations, the minimum detectable effect size with the adjusted sample of 219 participants, accounting for repeated decisions (clustering within participants) with an intra-class correlation of 0.2, is approximately 0.30. This corresponds to a medium effect size, which is commonly used in social science experiments.
Percentage: This means the study is powered to detect a 30% difference in the probability of correcting a teammate's or their own throws between the Info and No Info treatments, or a difference of similar magnitude in the expected values of replacement throws.
|