Friday, 13 November 2020
Serious problems in the reproducibility of brain imaging results
As I’ve pointed out in previous posts, the results produced by brain-imaging technologies such as fMRI are subject to at least two limitations. First. they detect neural activation only indirectly, by monitoring blood flows in the brain. Second, the methods used to analyze such images are subject to many forms of bias. These limitations were confirmed in a troubling study by Tom Schonberg, Thomas Nichols and Russell Poldrack, entitled “Variability in the analysis of a single neuroimaging data set by many teams ”, published in the May 20, 2020 issue of the journal Nature. (more…)
From the Simple to the Complex, From Thought to Language | Comments Closed