<p>Phuriku:</p>
<p>You’re talking about AP/Gallup polls. They are opinion polls. There are no right or wrong answers to opinion polls and they are not subject to the same tests that we’re trying to establish here. We need to come up with a reliable data set from which we can reliably extrapolate conclusions. We do not have that yet.</p>
<p>Moreover, you cannot reasonably argue that CC posters are ethnically representative of the UChicago student body at large. Without this baseline, all conclusions are questionable.</p>
<p>Moreover, you hit the nail on the head with your phrase, “significant change”. You simply cannot argue that there has been “significant change” based on the anecdotal and “common sense” arguments that you have presented. Imagine trying to argue that temperature differences between September 15, 2005 and September 15, 2006 are somehow statistically relevant. They are not. There is simply too much random and standard deviation to draw any significant conclusions from a data set that small. You would have to look at a much, much longer period of time to draw any significant conclusions on weather and temperature trends.</p>