MSNBC Distorts its Own Poll

Occasionally the MSM slips up and provides concrete proof of bias that is just too obvious to ignore.

Wednesday night after the Republican debate, MSNBC conducted an internet straw poll to gauge the results. On its web site under the heading "First Read" the question was asked "Who do you think won the Republican debate at the Reagan library?"

The poll results were presented in the form of a bar chart displayed with the accompanying data alongside. The results were clear: Ron Paul was the winner, and by a big margin.

How big? By more than double -- that's right, Mr. "Unelectable" beat Mitt Romney by 43.5% to 21.5%. The next runner up to Ron Paul was Rick Perry, with 16.4%.

But if you looked at the accompanying bar chart on the MSNBC web site showing the relative magnitude of votes for the candidates in the polling, it showed what almost looked like a photo finish at Aqueduct raceway - nearly too close to call.

Now a lot of folks will give the benefit of the doubt to MSNBC and say that at least it had the correct numbers in the chart, so what's the big deal? The big deal is that by nature humans are visual creatures more than cognitive ones. We tend to perceive visual relationships much more than numerical ones; numerical relationships require thought while visual ones are apparent at first glance and are persistent.

What MSNBC did was to present the viewer with the visual proposition that although Ron Paul won the debate, he didn't win by that much, so why get excited about it?

The fact is, the margin of Ron Paul's victory in the MSNBC poll is huge. Even if, as some might say, the polling was flooded by Ron Paul supporters, isn't that the point? These folks will presumably show up at the real polls and cast real votes.

When the intentionally misleading bar chart is corrected to show the true relationships, the difference is striking. Ron Paul won big -- and very big.

Below are shown the intentionally distorted chart and the visually corrected chart for comparison. The reader can decide how accurate MSNBC was in portraying its own poll results from the Republican debate.

Statistics may lie, but graphics do so far more effectively, and this has been going on for as long as there have been liars and graphics. The definitive text on this deceit is "The Visual Display of Quantitative Information" by Professor Edward Tufte. Acutely aware of the damage to social discourse caused by such misinformation, he writes:

"Lying graphics cheapen the graphical art everywhere...When a chart on television lies, it lies tens of millions of times over; when a New York Times chart lies, it lies 900,00 times over to a great many important and influential readers."

Haven't we had enough of this?

If you experience technical problems, please write to