Steve McIntyre: Anatomy Of The Lewandowsky Scam
The recent article by Lewandowsky et al, “NASA faked the moon landing, Therefore (Climate) Science is a Hoax: An Anatomy of the Motivated Rejection of Science”, has attracted commentary at Jo Nova, Bishop Hill, WUWT, Lucia’s and Skeptical Science, as well as by Lewandowsky himself. The data was placed online at Bishop Hill here.
Tom Curtis of Skeptical Science observed that at least “10 of the respondents have a significant probability of being produced by people attempting to scam the survey” and that the “paper has no data worth interpreting with regard to conspiracy theory ideation”. The scam identified by Curtis was that these respondents were almost certainly warmists caricaturing sceptics. Their caricatures were grotesque: two respondents, claiming to be skeptics, purported to believe in every single conspiracy, no matter how wacko. That the US government was complicit in 9/11; that the Moon Landings were fake; that SARS and AIDS were government plots, etc etc.
As others have observed, the number of actual respondents purporting to believe in the various conspiracies was, in many cases, very small. Only 10 respondents purported to believe in Curtis’ signature Moon Landing conspiracy. These included a disproportionate number of scam responses. Indeed, probably all of these responses were scams.
However, Lewandowsky’s statistical analysis was unequal to the very low hurdle of identifying these scam responses. Lewandowsky applied a technique closely related to principal components to scam and non-scam data alike, homogenizing them into a conspiratorial ideation.
In today’s post, I’ll extend Curtis’ analysis, identifying another “tell” that, in my opinion, can be used to identify scam responses that are not as overtly in-your-face grotesque as the scams already identified by Curtis. I’ll also consider whether the scam responses say something about the psychology of the scammers.