How does the Hot Air Survey match up to the pros? You decide.
posted at 7:15 pm on November 17, 2011 by Patrick Ishmael
A special thanks to Lori Williams at Tableau for her instrumental help in putting this together. I’ve used Tableau Public (which is free software) for a while to map data; this is the first time I’ve really had it graph my data. As you can see, it turned out great.
What you see below is a comparison between our long-running Hot Air surveys and a rolling average of professionally conducted national primary polls, as recorded by Real Clear Politics. Our polls are far from perfect, but I’ve long believed that if weighted properly, the Hot Air survey would stack up pretty well with the big guys at the big polling firms. I think these graphs generally confirm that theory; spikes in the professional polls coincide with spikes at Hot Air, and in some cases it’s even clear to what extent candidates are over or under represented among Hot Air readers (especially striking in the Romney slide.) If there are any stats nerds that have any insight into how to weight the data to get more “scientific,” I’d be much obliged. Regardless, our raw data has been really interesting and will always be a part of this project.