Here’s the notice for the paper, “Variation in chimpanzee ‘culture’ is predicted by local ecology, not geography:”
Shortly after our above paper was published in Biology Letters, we discovered several coding errors in the dataset we analysed. After re-analysing a corrected dataset, we did not find the same results as in our publication. In contrast, we found that no ecological variable was a statistically significant predictor of behavioural variation. Consequently, we do not feel that the main result of our publication is valid and have requested retraction of this manuscript.
So how did the errors come to light? Corresponding author Jason Kamilar, formerly of Yale and now at Midwestern University in Glendale, Arizona, tells Retraction Watch:
A colleague contacted me about a week after the paper appeared on Biology Letters’ Early View to request the dataset. We originally used a couple of different methods to code the data and we conducted the analyses about 2 years ago. After looking at the datasets in more detail I noticed several errors that likely was likely due to coding the data from a separately coded dataset instead of the original dataset. I confirmed the problem by re-analyzing the correct dataset and I obtained different results. I contacted the editors of Biology Letters and we agreed that retracting the paper was the best course of action.
We wanted to know whether the retraction would have a significant effect on the field.
I think the impact will be quite minor. The retraction occurred less than 2 months after the paper appeared online, and it was never actually published in an issue. In addition, I submitted a new version of the manuscript that contained the correct dataset and analysis, which is now in press in Journal of Human Evolution.
Not surprisingly, the new paper found the opposite of the original results:
…geography, and longitude in particular, was the best predictor of behavioral variation.
The authors were also transparent in the new paper, which includes a line noting the retraction:
Our paper also serves as correction to our recently retracted study (Kamilar and Marshack, 2011), which contained several coding errors in the dataset.
Kamilar, we should note, is as critical of others’ work as he is of his own. Late last year, he was a co-author of a Comment in Science alleging flaws in a May 2011 report on whether dinosaurs were nocturnal. Its abstract:
Schmitz and Motani (Reports, 6 May 2011, p. 705) claimed to definitively reconstruct activity patterns of Mesozoic archosaurs using the anatomy of the orbit and scleral ring. However, we find serious flaws in the data, methods, and interpretations of this study. Accordingly, it is not yet possible to reconstruct the activity patterns of most fossil archosaurs with a high degree of confidence.
The response from the original paper’s authors wasn’t anything like a retraction; it was more like doubling down:
Hall et al. claim that it is not yet possible to infer the diel activity patterns of fossil archosaurs with high confidence. We demonstrate here that this assertion is founded on unscreened data, untenable assumptions, and inappropriate methods. Our approach follows ecomorphological and phylogenetic principles in a probabilistic framework, resulting in statistically well-supported reconstructions of diel activity patterns in Mesozoic archosaurs.
See no evil, hear no evil?
Hat tip: Michael Balter