Data
We collected data from multiple sources to ensure coverage throughout the study period and to enable triangulation across different types of data (Gibbert et al., 2008). Data sources include field participation and observation, OSMS documents and survey reports, secondary documentation for each case, and interviews. We began observing and discussing how makers and maker spaces responded to the pandemic in March 2020. Over the coming months, members of our team participated in virtual conferences, maker space workgroups, and OSMS activities (Cavalcanti et al., 2021). See Appendix B for additional details about our data collection.
We collected archival data in the form of secondary documents to inform our understanding of how self-organizing activities evolved throughout 2020. Types of documents collected include news and blog articles containing interviews and timelines, websites, summary reports containing stories and statistics, instructions about PPE designs and volunteer procedures, workflow process documents and tools such as web forms and spreadsheets, governance documents such as policies and meeting minutes, and crowdfunding campaigns with chronological updates. We also compiled lists of organizations involved with PPE efforts in each city based on the OSMS roster of local response groups, web and media searches, and with input and vetting from initial contacts who were key organizers in each case. These lists provided an important starting point from which we developed our understanding of networks nodes and formation prior to beginning our interviews. Table 1 summarizes the data collection according to source and case study.
———————————————–
INSERT TABLE 1 ABOUT HERE
———————————————–
We developed an extensive protocol to conduct structured interviews remotely using Zoom to screen-share our instrument in Qualtrics. Our questions were informed by social network theory (new and existing ties, individual and organizational interactions, key events, etc.) and the entrepreneurial crisis response literature (compassion organizing, response effectiveness, etc.). We piloted the interview instrument with four subject matter experts in North American cities outside our sample to refine the protocol while also gaining additional understanding of the general phenomenon. Pilot interviews lasted 109 minutes on average.
We recruited informants across different organizations beginning with a few primary contacts who acted as key organizers in each city, then identified potential new informants using a combination of snowballing and emailing other contacts on our organization lists. Informants represented PPE producers, PPE recipients such as clinicians, as well as some people who acted in multiple roles or brokered contacts between people in different roles. Overall, we conducted 36 interviews with 35 informants across the four cases. Interviews occurred between December 10, 2020 and March 13, 2021 and lasted an average of 83.3 minutes. We conducted 31 structured interviews using the Qualtrics instrument and five shorter semi-structured interviews when that instrument was not feasible given time constraints for the informant. Interviews were recorded and transcribed using Otter.ai for a total of 1290 pages of transcriptions.