[Paper Review] Matching or Crashing? Personality-based Team Formation in Crowdsourcing Environments

Ioanna Lykourentzou, Angeliki Antoniou+1 more|arXiv (Cornell University)|Jan 26, 2015

Mobile Crowdsensing and Crowdsourcing38 references3 citations

TL;DR

This study investigates whether personality-based team formation improves performance in cooperative crowdsourcing. Using the DISC personality framework, it compares 'matching' (compatible) and 'crashing' (clashing) teams across collaborative and competitive tasks, finding that matching teams achieve significantly higher quality outcomes, better communication, lower stress, and greater satisfaction—especially in collaborative settings.

ABSTRACT

"Does placing workers together based on their personality give better performance results in cooperative crowdsourcing settings, compared to non-personality based crowd team formation?" In this work we examine the impact of personality compatibility on the effectiveness of crowdsourced team work. Using a personality-based group dynamics approach, we examine two main types of personality combinations (matching and crashing) on two main types of tasks (collaborative and competitive). Our experimental results show that personality compatibility significantly affects the quality of the team's final outcome, the quality of interactions and the emotions experienced by the team members. The present study is the first to examine the effect of personality over team result in crowdsourcing settings, and it has practical implications for the better design of crowdsourced team work.

Motivation & Objective

• To examine whether personality compatibility affects team performance in cooperative crowdsourcing environments.
• To compare the impact of 'matching' (compatible) versus 'crashing' (clashing) personality combinations on team outcomes.
• To evaluate how personality compatibility influences communication quality, stress levels, and team satisfaction.
• To assess the practical feasibility of using personality tests (DISC) for team formation in crowdsourcing platforms.
• To explore the implications of personality-based matching for improving quality and efficiency in crowdsourced group work.

Proposed method

• Employed the DISC personality model (Dominance, Influence, Steadiness, Compliance) to categorize participants into personality types.
• Formed four experimental team conditions: matching-collaborative, crashing-collaborative, matching-competitive, and crashing-competitive.
• Conducted a controlled experiment with 297 participants assigned to teams to develop advertisements under collaborative or competitive task structures.
• Measured outcomes using quantitative metrics: final result quality, communication quality, stress levels, and self-reported team efficiency.
• Applied statistical analysis (ANOVA and post-hoc tests) to evaluate differences across team types.
• Used ethical protocols, including informed consent and data anonymization, and collected personality data via a brief online test.

Experimental results

Research questions

• RQ1 Does personality compatibility significantly affect the quality of the final team outcome in cooperative crowdsourcing?
• RQ2 How does personality compatibility influence the quality of team interactions and communication?
• RQ3 What is the impact of personality-based team formation on team members' stress levels and emotional experiences?
• RQ4 Are there differences in team performance between collaborative and competitive task types when personality compatibility is considered?
• RQ5 To what extent does personality-based matching improve team satisfaction and perceived efficiency?

Key findings

• Teams with matching personalities produced significantly higher-quality final outcomes compared to crashing teams, especially in collaborative tasks.
• Matching teams reported significantly better communication quality and lower stress levels than crashing teams.
• Participants in matching teams, particularly in collaborative settings, reported higher satisfaction with the end result and greater self-perception of team efficiency.
• The probability of forming a crashing team through random selection is high due to the overrepresentation of dominant (D) and influential (I) personality types in the crowd worker population.
• Personality-based matching was found to be a strong intrinsic motivator, contributing to improved performance and well-being in team settings.
• The DISC framework demonstrated strong prognostic validity for predicting team dynamics, with high convergence to the MBTI personality model.