Lewandowsky's Psychological Science publishing hoax reaches the media

From Paul Matthew’s Climate Scepticism blog:

In the blogosphere, the hoax paper by Lewandowsky, Oberauer and Gignac,  NASA faked the moon landing—therefore, (climate) science is a hoax was exposed soon after it came out  –  see blog posts by Jo Nova, Steve McIntyre, Brandon Shollenberger, José Duarte and many others. Duarte’s comments are significant as he is a published researcher in the field, referring to the paper as a fraud and calling for its retraction.

The most concise explanation of the error/fraud is given at Kevin Marshall’s ManicBeancounter blog. The paper claims that belief in conspiracy theories predicts climate scepticism, yet of over 1000 in the sample, only 10 believed the moon landing was faked (rendering any statistical results completely meaningless) and of these 10 only 3 were sceptical about climate change. Despite the fact that the main claim of the paper and its title can be demolished in one sentence after simply looking at a table of the data, the journal concerned, Psychological Science, has taken no action.

Yesterday this story appeared in the Sydney Morning Herald. As far as I am aware this is the first time it has appeared in the mainstream media. Kudos to their reporter Paul Sheehan, and also to the Quillette magazine for writing about it in December, and Lee Jussim for giving a talk about it at an academic meeting on social psychology.  Here is an excerpt:

Distorted universities need a reality check

Cultural sensitivity is turning into a victory for ideology over objectivity

… universities have become havens for intolerance, orthodoxy and unscholarly distortion.

My favourite example, which encapsulates all of the above, was provided by Dr Lee Jussim, a professor of social psychology at Rutgers University in the US. He dissected a paper published by a respected journal,Psychological Science, in 2013, and found that it was rubbish, and probably published because the journal’s editors shared the ideological bias of the article’s conclusion.

The paper was entitled “NASA faked the moon landing – therefore (climate) science is a hoax“. The abstract of the study states: “Endorsement of a cluster of conspiracy theories (e.g., that the CIA killed Martin Luther King or that NASA faked the moon landing) predicts rejection of climate science … This provides confirmation of previous suggestions that conspiracist ideation contributes to the rejection of science.”

Note the term “conspiracist ideation”. The English language is being brutalised in the social sciences to create a false sense of rigour.

When Jussim checked the data, he found that of the 1145 participants in the study, only 10 thought the moon landing was a hoax. Of those who thought climate science was a hoax, almost all of them, 97.8 per cent, did NOT think the moon landing was a hoax.

The social psychologists who conducted the study had disguised the data and smothered it under a layer of obfuscation. No peer reviewer or journal editor took the time to check the raw data. Instead, the paper was published because it buttressed a pervasive ideological bias in the field…

This may provide some further embarrassment for the journal and its editors, particularly the remark made twice that the paper was published without questioning the data because its results fitted so well with their existing prejudices.

The editor of the journal at the time, Eric Eich, behaved fairly disreputably, as revealed by FOI correspondence, by responding to a question from Stephen McIntyre by getting Lewandowsky to respond and then including that response in his reply to McIntyre and declaring the matter closed (on a separate aspect related to ethics clearance).

The new editor, Stephen Lindsay, has recently written an editorial for the journal, Replication in Psychological Science, a subject that has been in the news recently.  He shows some understanding of the source of the statistical error discussed above, since there is a section on interpreting correlations in which he says it’s a good idea to draw scatter plots, but he doesn’t seem to understand the mechanism of the Lewandowsky hoax (find lots of people who think A and B are true, calculate a correlation, then falsely conclude that people who don’t think A also don’t think B, nicely illustrated by this plot of Lewandowsky’s data by Brandon).  Lindsay’s editorial article seems to lack real conviction – he just says he “encourages” authors to provide scatter plots. Worse still, there is no requirement to produce the raw data, he merely praises his predecessor Eich for instituting a system where authors who do provide raw data are rewarded with “badges”.  If these people were serious about tightening up their field, it would be a condition of acceptance that all raw data was produced before publication and not just made available but attached to the paper in some way.

At the end of his editorial Lindsay writes “If you have other ideas as to how to enhance Psychological Science, please e-mail me about them”. I’ve done that and drawn his attention to the Sydney Morning Herald piece and this post.