52. Two treatments have a nonsignificant difference. What additional evidence would support a claim that their effects are practically equivalent?
MediumStatistics & experiments
Technical
Show model answer
Specify a meaningful equivalence range before analysis and use an appropriate equivalence test or interval-based decision rule. The uncertainty must be narrow enough to exclude effects outside that acceptable range.
A large interval spanning important benefit and harm is inconclusive even if it includes zero. Explain the practical tolerance, design assumptions and achievable precision rather than translating failure to reject a difference into proof of sameness.
What interviewers listen for
- Define a practical equivalence range
- Require adequate precision
- Separate inconclusive from equivalent