RECAPPING THE ERRORS IN CHOOSING THE STUDY SUBJECTS
Design errors (e.g., smoking survey)
- The target population is not well suited to the research question (high school vs. middle school)
- Accessible population does not sufficiently represent the target population (cluster sampling)
- The intended sample does not sufficiently represent the accessible population because of biased sampling design (volunteers)
Implementation errors
- Due to random sampling error(chance)
- Due to systematic sampling error (bias)
- Failure to make contact
- Subject refuses to participate