UPDATE: The structure of the workshop may be slightly different based on the change of format to an online engagement.
Welcome and overview by workshop organizers
Brief participant introductions + Position paper lightning talk
Short break
Group Exercise
Discussion
Wrap-up
Click here for a list of workshop participants.
Workshop participants will be assembled into groups and presented with a set of hypothetical research tensions involving social media research. They will be given a brief in which they will be asked to do the following and then discuss at the conclusion of the workshop:
Outline the tension(s) presented in the brief.
Provide a list of considerations related to the tension(s).
Propose how researchers might address these tension(s).
What open questions remain related to the tension(s) presented?
Below are examples of research tensions that will be presented to the workshop participants. All tensions will be emailed to participants prior to the workshop so they have time to consider the nuances and complexities of the proposed tensions.
Researchers plan to scrape public comments from online newspaper pages to predict election outcomes. They will aggregate their analysis to determine public sentiment. The researchers don’t plan to inform commenters, and they plan to collect potentially identifiable usernames. Scraping comments violates the newspaper’s terms of service.
Researchers plan to scrape profile photos, which are visible to any member of the service, from a dating site to build models that predict sexual preference or behavior. Researchers will not inform the dating site users, but they will not collect any identifying information and their photograph dataset will not be released publicly. Creating a fake profile, necessary to access the photos, violates the dating site’s terms of service.
Researchers plan to scrape public posts and interactions from Facebook to study group-level dynamics. They plan to collect informed consent from the original poster, but not those they interacted with, and they may collect identifying information. Scraping posts with permission of the original poster does not violate Facebook’s terms of service.
Researchers plan to scrape data from an open teen forum and combine it with scraped tweets to predict mental health conditions. The researchers will not inform forum users, and they may collect potentially identifying information. Scraping data violates neither the health forum nor Twitter’s terms of service.
Researchers plan to scrape Facebook and Twitter where people are self-identifying as COVID+ and cross referencing it with voter registrations to track political affiliation and other demographic information connected to public COVID+ individuals.