Navigating the Ethics of Internet Measurement: Researchers' Perspectives from a Case Study in the EU
By: Sahibzada Farhan Amin , Sana Athar , Anja Feldmann and more
Potential Business Impact:
Helps internet researchers make safer, ethical choices.
Internet measurement research is essential for understanding, improving, and securing Internet infrastructure. However, its methods often involve large-scale data collection and user observation, raising complex ethical questions. While recent research has identified ethical challenges in Internet measurement research and laid out best practices, little is known about how researchers actually make ethical decisions in their research practice. To understand how these practices take shape day-to-day from the perspective of Internet measurement researchers, we interviewed 16 researchers from an Internet measurement research group in the EU. Through thematic analysis, we find that researchers deal with five main ethical challenges: privacy and consent issues, the possibility of unintended harm, balancing transparency with security and accountability, uncertain ethical boundaries, and hurdles in the ethics review process. Researchers address these by lab testing, rate limiting, setting up clear communication channels, and relying heavily on mentors and colleagues for guidance. Researchers express that ethical requirements vary across institutions, jurisdictions and conferences, and ethics review boards often lack the technical knowledge to evaluate Internet measurement research. We also highlight the invisible labor of Internet measurement researchers and describe their ethics practices as craft knowledge, both of which are crucial in upholding responsible research practices in the Internet measurement community.
Similar Papers
REFLECTing SPERET: Measuring and Promoting Ethics and Privacy Reflexivity in Eye-Tracking Research
Human-Computer Interaction
Helps eye-tracking researchers think about privacy.
Measuring What Matters: Connecting AI Ethics Evaluations to System Attributes, Hazards, and Harms
Human-Computer Interaction
Helps AI systems avoid causing harm.
Ethical conundrums: Hacked data in the study of far-right violent extremism
Computers and Society
Researchers can ethically use hacked data for studies.