Incorrect Scaling Factor and Errors in Figures and P Value Thresholds | JAMA Neurology | JAMA Network
[Skip to Navigation]
Sign In
Views 646
Citations 0
July 2017

Incorrect Scaling Factor and Errors in Figures and P Value Thresholds

JAMA Neurol. 2017;74(7):873. doi:10.1001/jamaneurol.2017.0586

In the Original Investigation titled “Association of Insulin Resistance With Cerebral Glucose Uptake in Late Middle–Aged Adults at Risk for Alzheimer Disease,” published in the September 2015 issue of JAMA Neurology,1 errors occurred in the Abstract, Figures 2 and 3, and the Results section of the text. While working on a follow-up analysis of the data, the authors discovered errors in the statistical model used in the original analyses and report.1 These errors affected the results reported in panel A of Figure 2, but the association between insulin resistance and glucose metabolism remains. Also, where a single contiguous cluster was previously reported (P < .05, familywise error corrected), the effect is now localized to distinct clusters (P < .001), but the effect remains robust in the medial temporal lobe (at a threshold of P < .05, familywise error corrected). In addition, the x-axes in Figure 3 were inadvertently reversed. This article was corrected online.2

Willette  AA, Bendlin  BB, Starks  EJ,  et al.  Association of insulin resistance with cerebral glucose uptake in late middle–aged adults at risk for Alzheimer disease.  JAMA Neurol. 2015;72(9):1013-1020.PubMedGoogle ScholarCrossref
Bendlin  BB.  Incorrect scaling factor and errors in figures and P value thresholds  [published online May 1, 2017].  JAMA Neurol. doi:10.1001/jamaneurol.2017.0509Google Scholar