Arbitron's sampling methodology is FAR more important than MrorMrsDePetro's misdeeds

[Brian Vita]

> If a sample can legitimately include six people from one 
> apartment building, why not get all 100 from the same 
> building? Many multi-family buildings are plenty large enough 
> to have more than 100 residents. 

The numbers would be even more significantly skewed if the "households" were
in an "ethnic" neighborhood - ie. Chinatown, the North End, etc

Brian