Steve McIntyre audited Stephan Lewandowsky’s data to weed out the obvious fake responses. That people would “game” the test was predictable given the clumsy nature of the survey, the one-sided nature of the conspiracies investigated, the virulently anti-skeptic sites where it was hosted, and the comments on the threads where it was announced. Obviously the survey hoped to show skeptics were nutters, and when it was posted in front of those who-hate-skeptics, readers obliged.
Steve McIntyre weighs in with a lengthy post, several original graphs, and concludes:
“Lewandowsky, like Gleick, probably fancies himself a hero of the Cause. But ironically. Lewandowsky’s paper will stand only as a landmark of junk science – fake results from faked responses.
As Tom Curtis observed, Lewandowsky has no moral alternative but to withdraw his paper.”
When the number of responses to conspiracies are graphed against the share that is “skeptical” of man-made global warming McIntyre reveals an interesting pattern. The “Oklahoma” point on the bottom right of the graph was the most popular conspiracy theory — but percentage-wise, “alarmists” were more likely to support this theory than so called “skeptics” were.
The line across the graph represents the proportion of the total responses which [...]