Skip to interview prep
Apexed

6. Measurements will be taken on different days with different operators. How could blocking improve the experiment?

MediumResearch questions & experimental design

Technical
Show model answer

Group runs around important known nuisance factors, such as day or operator, and compare the conditions within those groups where feasible. This helps separate the factor of interest from predictable background variation.

Randomise appropriately within the design and retain the block information for analysis. Do not put every control on one day and every treatment on another, because the treatment effect would then be entangled with the day effect.

What interviewers listen for

  • Identify known nuisance factors
  • Compare within suitable blocks
  • Avoid treatment-day confounding
Measurements will be taken on different days with different operators. How could blocking improve the experiment — Science & research interview question — Apexe