Comment & Response
March 2014

The Problem of Spurious Correlations Between Pairs of Brain Metabolite Values Measured in the Same Voxel With Magnetic Resonance Spectroscopy—Reply

Author Affiliations
  • 1Department of Psychiatry and Behavioral Neurobiology, University of Alabama at Birmingham
  • 2Department of Biostatistics, University of Alabama at Birmingham

Copyright 2014 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.

JAMA Psychiatry. 2014;71(3):339-340. doi:10.1001/jamapsychiatry.2014.22

In Reply Dr Maddock’s letter in response to our recent article1 suggests that it is statistically problematic to report correlations between ratios without accounting for spurious correlations. Spurious correlations were first described by Pearson, who showed that correlations between 2 ratio variables that have the same denominator may be present, even when the original variables from which the ratios were derived are uncorrelated.2

First Page Preview View Large
First page PDF preview
First page PDF preview