53. All samples in one biological group were processed in one batch and the comparison group in another. Why is interpretation difficult?
HardLife & environmental research
Technical
Show model answer
Group and batch are confounded, so a measured difference may reflect the biological comparison, processing conditions or both. A statistical adjustment cannot automatically recover information that the design never separated.
Inspect controls and available evidence, limit conclusions and plan a suitable balanced or independent follow-up if feasible. Do not promise that normalisation will prove the biological explanation merely because the adjusted plot appears cleaner.
What interviewers listen for
- Recognise design confounding
- Do not overpromise computational correction
- Seek independent discriminating evidence