Ethical frameworks and philosophy

• Kant - what we ought to do - what’s moral
  • moral rules
  • use of data violates our autonomy
• utilitarianism: overall happiness (sum total)
  • how do we measure this for big data?
• social contracts: we have “agreed” that to be part of society means we follow certain rules
  • have we agreed to share our data in exchange for services?
• virtue ethics: certain things are virtuous / morally correct
• unalienable rights (eg to your data)
• individual rights

Neutrality - data

• depends on how it’s used
  • issues of confirmation bias - trying to use data to fit what we think
• the data itself is neutral
• depends on how we collect the data - there’s more data on some people than others
• depends on the size of the dataset (small data may not be accurate)

Neutrality - algorithms / programs

• because it’s complicated you can’t determine its bias
• not transparent
• what’s included in the program / algorithm affects the result
• machine learning algorithms can skew based on available information
• important to think about the outcome

Ethics of the census data

• How is information used?
  • Determine representation in congress
  • Determine funding
• Concerns about the citizenship question - what if data is used for harm
  • but it’s important to know who is living where

Neutrality of census data

• information is only available to those who need it / the census bureau
• because of immigrant concerns people may lie or not fill out the census and that skews the data