This study explored crowdsource volunteer systems using data collected from the Anti-Pandemic Pioneers (a.k.a. Pioneers), a mobile platform for organizing community-level volunteer activities in Shenzhen, China. Launched at the start of the COVID-19 epidemic in February 2020, the Pioneers platform allows users to organize and take part in local volunteer activities, and was quickly adopted by the community staff and volunteers in Shenzhen, with more than 679 million volunteer participations in two years. By analyzing volunteers' collective behavior, the team revealed the impact of epidemic situations, policy, and platform mechanisms on volunteers' self-organization behavior and gave suggestions on improving the organizational effectiveness in crowdsourced volunteer systems.
In social sciences, self-organization refers to the process that occurs when disordered systems spontaneously attain order through simple interactions among individuals. The research team found that self-organization also exists in volunteer activities, as volunteers adapt to changing environments through rapid self-organized actions. For example, new volunteers are often more inclined to explore different volunteer tasks and groups. As they gain experience, their task selection patterns will gradually stabilize until a change in community needs (external environment) causes uncertainty in their behaviors. The present study proposed a quantitative way to measure self-organization in volunteer behaviors based on normalized conditional entropy (Figure 1). Using this tool, the study analyzed how self-organization in volunteers' behaviors regarding task participation, task organizer preference, and task preference evolve from 2020 to 2021, revealing how communities in various districts of Shenzhen respond to different types of community needs.
Figure 1. Overview of the volunteer behavior model
Causality analysis is conducted using the PCMCI+ algorithm among a variety of internal factors (e.g., number of volunteers, number of task organizers, etc.) and external factors (e.g., epidemic and anti-epidemic policies, holidays, and policy changes in the platform) related to volunteers' self-organization (Figure 2). Results showed that the daily number of new COVID-19 cases had an indirect causal effect on volunteers’ self-organization by mediating its impact through the number of volunteers and the distribution of task categories. Meanwhile, centralized interventions significantly influenced volunteers' participation rate and their preferences for organizers and tasks.
Figure 2. Two causality graphs for organizer preference NCE (O-NCE).
To further investigate the link between volunteer self-organization and community management effectiveness, the team simulated the task participation process of volunteers using agent-based modeling and compared the organizational strength and task completion efficiency of three volunteer organization schemes (self-organized, platform-centralized, and hybrid) (Figure 3. a-d). The experiments demonstrated that the self-organized scheme is most adaptive to uncertain community demand; the centralized scheme is most efficient when the demand is known, and the hybrid scheme balances both adaptability and organizational efficiency. As a result, the research team suggests that early top-down leadership from the platform combined with bottom-up self-organization efforts in the community can create the most resilient societal governance force against public crises such as the COVID-19 pandemic.
Figure 3. Schematic diagram and experimental results of agent-based simulation.
This research was recently published in the journal Humanities and Social Sciences Communications entitled Optimizing self-organized volunteer efforts in response to the COVID-19 pandemic. The corresponding author is Prof. Yang LI. The co-first authors are Anping ZHANG and Ke ZHANG, Ph.D candidates at Tsinghua SIGS. Authors also include Wanda LI, Prof. Yue WANG, and Prof. Lin ZHANG.
Link to full article:
https://www.nature.com/articles/s41599-022-01127-2
Written by Anping Zhang, Ke Zhang,Yang Li, Lin Zhang & Yue Wang
Edited by Alena Shish & Yuan Yang
Images by Anping Zhang, Wanda Li & Ke Zhang
