I didn’t contribute to the Obama campaign and have not received an H1N1 flu shot. Does this mean I break the statistical analysis, being 0 for 0? Or does it mean that my anecdotal evidence is merely irrelevant.

I have a young nephew and a young niece who have received the shots, but I’m pretty sure that neither of them contributed to the Obama campaign. (At 4 and 1.5 years old, their discretionary spending is limited.) I wonder if they also break the statistical analysis.

But as I said, tough questions. I suspect, to have a fair shot at some kind of useful statistical analysis, rights to privacy in the case of #1 would limit the sampling size to an unusably small sampling. Of course this doesn’t preclude the use of statistics to prove whatever needs proving; so I’d say make up answers, which is about as good. No one else will have a sampling size that could lead to a refutation of your statistics, anyway.

#2 is a bit different, since certain public corporations and institutions probably need to keep track of, and report, usable information. OTOH, private organizations probably don’t need to report such things, so we default to the same difficulty we have in answer #1 statistically. Again, however, you can focus only on certain organization who do report the necessary information and, as is now accepted practice, use your results to create a universal statement of fact covering also the private institutions and organizations. I, for one, would like to know whether the 5GW organizations and various other secret societies in America have received a benefit from supporting Obama, so a study building a universal assessment upon a very small sampling of publicly traded corporations would come in handy.