Finance Analysis
Some people sometimes stratify a randomization on one or more variables. Under stratification, researchers try to make sure certain attributes are balanced across the treatment and control groups. They stratify based on attributes that are most likely to have a strong influence on the outcome of interest.
1. Why is it good idea to stratify out sample by variables that are likely to have a strong influence on the outcome? For instance, why might we want to stratify on gym membership if we are studying the effect of an incentive to exercise?
2. Is stratification for this reason more important in small or large samples? why?