4. What does random assignment protect against, and what does it not guarantee?
EasyResearch questions & experimental design
Technical
Show model answer
Random assignment reduces systematic allocation differences between comparison groups, including influences the researcher may not know to measure. Its implementation should prevent predictable or discretionary reassignment.
It does not guarantee identical groups in a finite sample, correct measurement or a representative population. Check allocation procedures and analyse the design appropriately instead of treating the word randomised as proof that every source of bias disappeared.
What interviewers listen for
- Reduce systematic allocation bias
- Protect implementation
- Randomisation is not universal validity