Why this is mistaken
Separate one-sample intervals estimate separate parameters. Their overlap is not the sampling distribution of the difference. Describe the samples and request an appropriate two-sample design and method before making that claim.
Where to watch
When this mistake may appear
- Two groups each have a confidence interval.
- A chart invites a visual significance judgment.
Check your work
Your work may contain this mistake if:
- Declares different population means solely because intervals do not overlap.
- Declares equality solely because intervals overlap.