I've measured ADC values of liver and spleen in 50 patients, and these patients underwent three rounds of DWI at different time points (so there's total of 150 ADC values measured each for liver and spleen). What I want to do is compare the variability of liver ADC and spleen ADC at different time points. At first, I calculated coefficient of variance (CV) on the basis of three data points per patient, and then compared the CV of liver and CV of spleen using Wilcoxon test. But the reviewer recommended the following method:
"by dividing each individual data point by the mean over the entire sample, they derive measures having a standard deviation equal to the CV of the original data. Likelihood ratio tests or F tests in a repeated measures or mixed model regression analysis could then be used to compare the variances of the rescaled data, which is equivalent to comparing the CV values of the unscaled data."
Can anyone give me some more detailed information about this recommendation?
How can I compare the variability of ADC values at three different time points?