Closing the Psychological Distance: The Impact of Social Interactions on Team Performance - An Experiment

Pairs of university student participants, who are not acquainted with each other, are formed and then randomly assigned to either the Social Interaction group (intervention group) or the No Social Interaction group (control group). In the intervention group, participants discuss work-related topics (e.g., "What do you think are the essential elements for the success of team activities?" "What elements could be obstacles in carrying out team activities?" "What are the crucial qualities that an ideal leader for the team should possess?") for 10 minutes with their teammate and record their own opinions on their provided response sheets. In the control group, participants individually perform the same task as the intervention group, without discussing with their teammates, and write down their opinions on their response sheets. Thus, engaging in discussions within the team, exchanging, and sharing opinions constitute the intervention in this experiment.

2024-04-19

2024-04-20

1. Definition of Primary Outcome Measures:
- Prosociality Level: We will assess the degree of prosociality each team member displays towards their teammate by evaluating the level of affinity they feel towards their teammate and their motivation for the subsequent SLAP task, focusing on whether the motivation is for their own benefit or for their teammate's benefit. This assessment will be conducted using a Visual Analogue Scale (VAS)-like method, where participants will mark a vertical line on a 10 cm line segment that best represents their feelings.
- Prosociality Level Gap: We will measure the difference in prosociality levels between pairs within each team.
- SLAP Task Performance Metrics: The number of correctly typed words within the time limit (score), accuracy, and the rate of score improvement across the three sessions will serve as performance metrics. We will compare these metrics' averages between the two groups.

2. Analysis Methods:
- Statistical Methods: To compare the means of the prosociality levels and SLAP task performance metrics, we will use independent samples t-tests and multiple regression analysis. The latter will measure the effect of the intervention dummy, controlling for team attributes, gender composition, and members' personality traits as explanatory variables.
- Data Preprocessing: Samples from the control group where participants engaged in conversation or exchanged information during the 10-minute individual task, as well as samples where participants cheated or dropped out during the SLAP task, will be treated as missing data.

3. Hypotheses:
- We hypothesize that social interaction within teams will reduce the prosociality gap among team members. Specifically, we predict that the intervention group will exhibit a smaller prosociality gap within teams compared to the control group.
- Furthermore, we hypothesize that social interaction will enhance team performance in the SLAP task. We predict that the intervention group will achieve higher performance in the SLAP task compared to the control group.
- This hypothesis aligns with the theoretical analysis from Hattori and Yamada (2023), "Closing the Psychological Distance: The Effect of Social Interactions on Team Performance", MPRA Paper No. 117042, and aims to provide empirical support for their theoretical hypothesis.

4. Timing:
- The assessment of prosociality will be conducted immediately after the 10-minute team (or individual) task.
- The SLAP task will be performed using a computer immediately after a 10-minute briefing about the task, following the prosociality assessment. The task's performance will be automatically recorded.

In our study, we will delve into several exploratory questions alongside our primary hypotheses to gain deeper insights into the dynamics of team interactions and performance. Specifically, we aim to:

1. Examine the Interactions Between Demographic and Personality Traits**: We will gather data on participants' gender, age, and a range of self-reported personality traits, including altruism, competitiveness, cooperativeness, reciprocity, risk preference, proactiveness in relationships, openness in relationships, leadership, and self-perception of role in team tasks. These traits will be analyzed as control variables and potential moderators to assess how they may influence the effectiveness of the social interaction intervention on team dynamics and performance outcomes.

2. Analyze Post-Task Team Dynamics: After the SLAP task, participants will provide feedback on various aspects of team atmosphere and dynamics, such as the extent and quality of team conversations, adaptability, task enjoyment, and team cohesion. We will use this data to perform a mediation analysis, aiming to uncover the underlying mechanisms by which the social interaction intervention might affect SLAP task performance. Additionally, we will explore the relationship between these team dynamics and the scores achieved in the SLAP task, potentially employing cluster analysis to identify distinct patterns of team characteristics and their correlation with performance.

These exploratory analyses are intended to supplement our primary findings, offering a broader understanding of the factors that contribute to team effectiveness and how social interactions can shape these outcomes.

1. Experimental Design:
A two-arm randomized controlled trial (RCT) to investigate the effects of social interaction within teams on prosociality levels and task performance.

2. Experimental Design Details:
Teams of two university students who do not know each other are formed. One group, the intervention group, engages in a 10-minute discussion on work-related topics, while the control group does not interact and contemplates the topics individually. Both groups then perform the SLAP task, and data on prosociality levels and task performance are collected and compared.

Participants are recruited by encouraging voluntary participation among students enrolled in the "Experimenting with Human Behavior" lecture conducted by the primary investigator at his university. The lecture is open to undergraduate students from first to fourth year across all departments of the university.

Participants not only receive partial course credit for their participation but also compete in the SLAP task within the experiment, where the top-performing team in both the intervention and control groups will receive ¥1,000 (equivalent to $6.75) per team member. The monetary incentive is designed to foster continuous motivation among teams, as it is based on the highest score achieved in one of the three sessions of the SLAP task, ensuring that teams remain motivated throughout the task.

For detailed information regarding the experimental procedures, the specific questions and response formats used, the content of the 10-minute team (or individual) task, details of the SLAP task, and screenshots, please refer to the `Supporting Document' section.

Participants are randomly assigned to either the 'Social Interaction' group or the 'No Social Interaction' group using a computer-generated randomization sequence.

The unit of randomization is the team, where each pair of participants is randomly assigned as a unit to either the intervention or control group.

No

The study plans to include 40 pairs in each group, with each pair consisting of two participants. The experiment will be conducted in two sessions with different pairings within each group, resulting in 40 unique pairs per group.

The total planned number of observations corresponds to 80 pairs, considering that each pair's performance on the SLAP task is treated as a single observation, with the experiment being repeated in two sessions.

- For the Social Interaction group, there are 40 pairs planned (80 participants, effectively 40 unique individuals per session).
- For the No Social Interaction group, there are 40 pairs planned (80 participants, effectively 40 unique individuals per session).

Given our study's design, where 80 participants are involved in a repeated-measures setup with two sessions, and each session includes 40 pairs per treatment arm, our sample size is fixed at 40 pairs per arm for each session. We aim to detect a large effect size (cohen’s d = 0.8), assuming a significance level of 0.05 and a power of 0.95. This decision is pragmatic, considering the absence of prior research or pilot studies to guide our effect size estimation. Our study is therefore designed to detect large effects, balancing the feasibility of participant recruitment with the objective of identifying significant effects within the logistical constraints of our research framework.