“Student Perceptions of Writing Instruction: Twitter as a Tool for Pedagogical Growth”
by Sarah Lonelodge and Katie Rieger | Xchanges 15.1, Spring 2020
Participants and Ethical Considerations
Information concerning participant data such as gender, sex, age, race, year in school, area of study, and other information was not collected. Because this information is difficult to gather via Twitter, these demographics were not functions within our analysis. Therefore, the data discussed involves only the text and visuals in the tweets.
In collecting data, we carefully considered the ethical implications of studying personal posts. To maintain the integrity of the study, we employed Heidi McKee and James Porter’s (2008) heuristic for internet-based research which categorizes these considerations based on the public or private nature of the site, use of identifiable data, interaction with participants, topic sensitivity, and subject vulnerability in order to determine whether informed consent is necessary. Most of these questions were answered by the nature of Twitter and the design of this study which involved no interaction with Twitter users. As a platform that is open to anyone with access to the internet, Twitter was created with the goal of helping users gain a large number of followers, retweets, and likes. Because users are aware tweets are public, we determined that informed consent was not necessary. All tweets were collected using Twitter’s search feature that only displays public posts. However, even though the tweets were public, we strove to maintain anonymity by removing any personal or key identifiers (name, location, etc.) as we coded. Because we did not collect personal information and because Twitter is a public website, we believe that our data collection practices align with the ethical standards described by Michael Zimmer (2010), who discusses research using personal information collected from Facebook, as well as by Casey Fiesler and Nicholas Proferes (2018), who discuss research ethics on Twitter. As suggested by these and other scholars, we have avoided mining profiles for personal information, anonymized tweets, eliminated identifying information, and avoided sharing sensitive information with others. Additionally, as suggested by Hibbin, Samuel, and Derrick (2018), we have tried to look at “the qualitative nature of risk within individual [tweets] to protect participants” (p. 9) while still facilitating research and have only quoted tweets when absolutely necessary for discussion purposes.