protection from harm obviously.
manipulation of results particularly because some participants may present data which is too explicit for publishing.
Consent.
Permission to use results?
etc etc
The personal experiences of the researcher might create a bias- if they are looking at qualitative research, they might interpret a response differently than the responder intended because of their own experiences.
Obviously not identifying respondants, as these would be very personal responses, I assume?
Parental consent is always important for minors!