they initially surveyed 1027 people, some % of them were AA. They then wanted to see if AA vs "white" had similar opinions as groups. They didn't have enough AA initially to be statistically accurate. They added a lot more AA to do so.
There were not 309, or anything close to it, AA in the initial survey.
They conducted 172 extra interviews. That means there were 137 AA's in the initial survey. Seems perfectly acceptable. (pg 3 of the pdf)
Getting to the point, why did "independents" poll similarly to the "lwnjs" ?
Finally, please READ before wasting my time in the future.
Last edited by SlevinKelevra; 02-22-15 at 10:53 AM.