Originally Posted by N/Apower
I don't see how I lost you except that you don't understand how I calculated percentages without biasing for sample-size. I outlined that proceedure for you above so you can check the numbers (I can't help if 1 or 2 more people answered, so it might be 1-2% off or something by now). Take the raw data out of the poll, calculate to negate sample-size disparity between the groups, and there you go. Easy as pi.
Correct, I misinterpeted what you were showing for percentages. But still, as you admitted to an extent, the data leaves much to be desired. Nice shot in the dark though. Maybe you can refine your polling to get more telling results.