What does a 'blind' study design primarily aim to minimize?
Sampling bias
Subjectivity or observer bias
Measurement error
Non-response bias