Internet and Research Ethics Ralph Schroeder 4th July 2016 SDP
Overview An introduction to ethics: Kant versus Mill Context specific examples Gathering data directly from individuals Analyzing interaction in virtual environments Conducting Big Data research What is new? Your ethical questions
An introduction to ethics Kant versus Mill Historical abuses
Gathering data directly from individuals Protection of harm More difficult to assess Strategies (make it clear participants can leave, prior rapport with participant, establishing netiquette) Ensuring anonymity and confidentiality Perceived anonymity of the Internet To be considered at all stages of the research
Gathering data directly from individuals Informed consent Distance between researcher & participant, challenges anonymity strategies, verifying ability to give informed consent Strategies (readability of documents, use of quizzes, recruitment strategy and verifying identity) Protected not burdened When to employ a human subjects model?
Analyzing interaction in virtual environments Focus on graphical online spaces with avatar interaction Differences in text-only versus voice communication, video- versus virtual, etc. Contexts of use include online gaming, spaces for socializing and collaborating, training online for offline tasks, experimenting in virtual
Research on VEs: Contexts and the role of the researcher Contact people offline? Weigh burden on research participant The online social setting: formally public, but respect the conventions for the privacy of the space? Be sensitive to context Disclose researcher identity? Online possibilities are different from offline (ID tag) ‘Invasion’ of researchers Respect social milieu
Research on VEs: data capture Tools for capture are more powerful than for capturing offline interactions Anonymous data about populations, but surveillance? Reproducing and anonymizing captured interactions, but possible identification by search?
Research in online worlds Maintaining the trust of environments and avatars and the persons ‘behind them’ New possibilities for the study of social interaction in online worlds and VEs Virtual Milgram as example A balance of deontological and utilitarian research ethics
Virtual and other online environments Environments for Being There Together (experiments) Online Networks (online to offline identification, surveillance) Ethnographic and participant observation Large-scale data capture, blurring of commercial boundaries From virtual to real behaviour modification
Data and Big Data Data Big data Belongs to object Comes before analysis The most divisible, atomizable unit Big data Greater scale and scope than previously available for given phenomenon
Emotional Contagion Study Almost 700,000 users, divided into two randomly selected groups 3 million posts and 122 million words were analysed Users were unaware of contagion and of study Unexpected result: Positive words in timeline lead to more positive words posted, and vice versa Kramer A, Guillory J and Hancock J (2014) Experimental evidence of massive-scale emotional contagion through social networks. Proceedings of the National Academy of Sciences. 111 (24): 8788-90.
Facebook ‘emotional contagion’ Three issues: research ethics, legal, and scientific ‘manipulation’ Research ethics IRB’s apply to academic research Legal issues are covered by ‘terms of service’ Scientific ‘manipulation’ is a product of big data Sources: Kramer, Adam; Guillory, Jamie and Hancock, Jeffrey. 2014. Experimental evidence of massive-scale emotional contagion through social networks. Proceedings of the National Academy of Sciences, 111 (24), 8788-90. http://laboratorium.net/archive/2014/06/30/the_facebook_emotional_manipulation_study_source
Ethical and Social Implications Research ethics, legality, and the larger issue of the increasing power of big data in society Regulation will ‘only drive research underground’ (research ethicists) - versus more regulation and transparency are needed Non-replicable findings (compare Google flu trends)
Uses and Limits Social science and scientific uses are limited by data sources Legal and ethical debates and regulations will continue Public concern will continue Populations will be increasingly subject to data analysis, subject to data and regulatory limits Orwell versus Huxley
What is new? Institutional vs. individual responsibility Challenges in devising a code of practice in a global context Disciplinary boundaries Less well known implications Characteristics of data Abundance, distance between creator of data and researcher, data as a “commodity” Differences between ethical and legal provisions
On ethical decision-making “At its most fundamental level, we recognise that ethical decision making interweaves one’s fundamental world view (ontology, epistemology, values, etc), one’s academic and political environment (purposes), one’s defining disciplinary assumptions, and one’s methodological stances. Decision making occurs at many junctures in the cycle of inquiry, including research design, research conduct and research production and dissemination” (AoiR, 2012:3).
Key considerations Human subjects vs humanities model of ethics Sensitivity to context & participants Private vs public Ethical standpoints can change throughout the research Protect - not burden - participants Take responsibility for ethical and research choices Do meaningful & high quality research Write about ethical decision making
Your ethical questions........& potential strategies