Why this is mistaken
Random selection addresses how units enter from the frame. It does not add missing units, remove duplicates, fix dates, align classifications, or correct the wrong population definition. Test sampling and nonsampling error separately.
Where to watch
When this mistake may appear
- A sample is described as random.
- A confidence interval is numerically narrow.
Check your work
Your work may contain this mistake if:
- Claims bias is impossible after random selection.
- Uses margin of error as an allowance for wrong source records.