Closed ejm714 closed 6 years ago
Add to framing:
Pithy, actionable
Not all have to be relevant to every project (N/A
is an ok answer)
It is ethical to be good at your job, but this is not a list of statistical best practices
Provoke discussion, not recommending particular fixes or policies
Don't create comprehensive lists of what is unfair or harmful, this is up to the teams that use the checklist
Have we noted where we are vulnerable to bias from missing perspectives?
The list follows the data science process and not all projects engage all the steps, so may not be relevant
ethics is hard, so these conversations may also be hard
Submit PR as a way to get the convo started (references give some ammunition)
Add to contributing:
Want to programmatically generate them so they both get updated when checklist or references are updated.