SICSS-Rutgers 2022

Day 1: Research Ethics

June 20th - July 1st, 2022 | Rutgers University - New Brunswick

We thank the core sponsors of the Summer Institutes in Computational Social Science, the Rutgers University- New Brunswick Library, and the School of Arts and Sciences at Rutgers University for their support of the event.

Ethics

Material Based on

• Ethics and Computational Social Science (part 1)
• Ethics and Computational Social Science (part 2)
• Ethics Additions and Extensions
• Bit by Bit Chapter 6

​

Ethics

Approaches

• Rules-based (following the IRB, common in social sciences)
• Concerns: slow to change, provides poor explanation for decisions
• Ad hoc approach (common in data science)
• Concerns: process difficult, time-consuming
• Bit by Bit suggested approach: Principles-based

​

Ethical Principles

Ethical Principles

• Respect for persons (i.e., informed consent)

​

• Beneficence (i.e., minimize risks, maximize benefits)

​

​

​

Example: Is Informed Consent Possible?

Ethical Principles

Ethical Principles

• Respect for persons (i.e., informed consent)
• Beneficence (i.e., minimize risks, maximize benefits)
• Justice (i.e., distribution of burdens and access to benefits)
• Respect for Law and Public Interest (i.e., comply with laws, be transparent)

Apply Ethical Frameworks to Balance Considerations

• Consequentialism (focus on ends)
• Deontology (focus on means)

Respect for Law: Web Scraping

DNC website

Library Data Archives (NYU Law)

Factiva News Archive

Example: Audit Studies

Example: Audit Studies

Example: Audit Studies

“It is extremely, extremely unlikely (that this is an audit study) given the low number of audit studies conducted with higher education samples. Emphasis on extremely.”

Some Academics start kvetching/lampooning on Twitter

Others defend audit studies

(e.g., Mike Kenwick complains to his friends on gchat, but remains silent on social media)

Example: Audit Studies

Ethical Challenges

Areas of Difficulty

• Informed consent (e.g., weighing potential harm to subjects vs. context + social benefits)
• Informational risk (e.g., exposure of sensitive data through re-identification)
• Privacy (e.g., public / private dichotomy vs. contextual integrity)
• Making decisions in the face of uncertainty (i.e., no risk-free approach)
• Unintended secondary uses

​

Ways Forward

• Examples: benchmarking risk to everyday life, staged trials, data protection plans
• View IRB as a floor (e.g., the SSRC has multiple guidelines for ethics review)
• Imagine how other people would feel if part of the study, observing the study
• View ethics as continuous
• Ethics as a research opportunity

YOUR TURN: GROUP ACTIVITY