RESEARCH ETHICS AND PUBLIC TRUST, PRECONDITIONS FOR CONTINUED GROWTH OF INTERNET MEDIATED RESEARCH (slides)

RESEARCH ETHICS AND PUBLIC TRUST, PRECONDITIONS FOR CONTINUED GROWTH OF

INTERNET MEDIATED RESEARCHPublic Confidence in Internet Mediate Research

Ansgar Koene, Elvira Perez, Christopher J. Carter, Ramona Statache, Svenja Adolphs, Claire

O’Malley, Tom Rodden, and Derek McAuley HORIZON Digital Economy Research,

University of Nottingham

• In the public eye – public interest in Internet Mediated Research

• Lessons from history – controversy and public backlash

• Ethics requirements– Public-private domain distinction– Confidentiality and Security– Valid informed consent– Withdrawal and debriefing– Scientific value and potential harm

• Ethics by design – Privacy protecting architectures• Conclusions

Overview

• For better of for worse, internet mediated research draws media attention

• Association with headline-leading companies like e.g. Google, Facebook

• Association with topic buzz-words, e.g. Big-Data• Promise of giving quantitative approach to social

science questions that are important for policy issues– Evidence based policy

In the public eye

• So far the public attention has resulted in– Digital Economy identified as priority area for future economic growth

– Increased research funding

• What could possibly go wrong?

In the public eye

News media love controversy!You are being manipulated!

News media love controversy!All your secrets are revealed!

Lessons from history: GM crops

In the 1990s, Bio-tech and Genetic Manipulation ruled headlines•1990: Human Genome Project started

•1994: First GM crop approved for sale in US

•1997: Dolly the sheep cloned

•1998: EU stops further approval of GMOs

Lessons from history: GM crops What went wrong?

• Loss of public trust in regulatory institutions, scientists and industry– Failure to address people’s fears about unintended/unpredictable side effects.

– Failure to address transformative nature of technology on society and social structures

– Failure to take due account of public concerns about the motives of those developing the technology and the regulatory frameworks.

• Dismissive attitude towards ‘irrational’ public concerns, focusing on educating people with facts and information.

Resulting policy changes• Responsible Research and Innovation (RRI) agenda

– Focus research and innovation to achieve social benefit

– Consistent, ongoing involvement of society, including public & NGOs

– Assess and prioritize social, ethical and environmental impact, risk and opportunities alongside technical and commercial

– Develop oversight mechanisms to anticipate and manage problems and opportunities

– Openness and transparency as integral components of research and innovation

Impact assessment exercises

• Code of Human Research Ethics (BPS 2013)1. Respect for the autonomy and dignity of persons2. Scientific value3. Social responsibility4. Maximising benefit and minimising harm

• Challanges for Internet Mediated Research1. Public-private online domain distinction2. Confidentiality and security of online data3. Procedures for obtaining valid informed consent4. Procedures for debriefing and right-to-withdraw5. Implications for scientific value and potential

harm

Ethics requirements and challenges for IMR

• Any data that was clearly derived fro, specific online communications will always be experienced as intimate and personal by the person who posted it.

• Legal distinction between public and private domains is irrelevant for personal experience of citizens.

• High level of trust in the researchers is required to make people feel comfortable with research on their data.

Public-private domain distinction

• Anonymization is one of the most basic steps for maintaining confidentiality, showing respect and gaining the trust of research participants.

• Anonymity is even more important when dealing with social media posts where access to the raw source data cannot be controlled by the researcher.

• Wealth of mineable secondary information sources means any hint at identity of participant can lead to de-anonymization.

Confidentiality and online security

Valid informed consent

• Core value to respect autonomy and dignity of persons.

• Probably the most frequently violated principle– unread, unintelligible terms-of-use

• Obtaining true valid consent from authors of social network posts is:– a technical and potentially labour intensive challenge

– goes a long way towards establishing a conscientious and trustworthy reputation.

• Participation provides deeper understanding of a study• Right to withdraw supports validity of consent• If the nature of the study requires that participant

must be naïve of the true purpose, debriefing and right to withdraw are crucial for valid consent.

• Challenges for research with remote interaction with participants:– How to know if participants paid attention to debriefing?

– How to facilitate right to withdraw without breaking anonymity

Withdrawal and debriefing

• Failure of dealing with these questions in a thorough and transparent manner produced the public backlash in the GM crops controversy.

• Controversies over social responsibility are threatening to become a recurring theme in the popular media:– “Facebook reveals news feed experiment to control emotions”

– “OKCupid experiments with ‘bad’ dating matches”

Scientific value and potential harm

• Encourage compliance with ethical research protocols by making it as easy as possible– provide tools that are intrinsically privacy-respecting.

• Especially important for IMR where much of the research is done by corporations or other groups that lack ethical review board structures.

• Three such tools are:– PRISONER (Privacy-Respecting Infras- tructure for Social Online Network Experimental Research) architecture (Hutton et al. 2012)

– Dataware system (Mortier et al. 2013)– The Ma3tch (autonomous anonymous analysis) technology (Kroon 2013)

Privacy protecting architectures

• Internet Mediated Research is receiving a greater level of media scrutiny than most other types of research.

• There are still many uncertainties around aspects of IMR ethics

• This combination raises the risk for triggering a controversy and public backlash similar to the GM crops in Europe in the 1990s.

• To avoid such a controversy it is essential to retain the confidence and trust of the public which depends heavily on the use of responsible safeguards for privacy and ethical treatment of human data.

Conclusion

Thank you

For more information on this topic, visit http://casma.wp.horizon.ac.uk/

The PRISONER architecture

Dataware

Ma3tch