NeurosciencebioRxivPreprint
Statistical significance in DTI group analyses: How the choice of the estimator can inflate effect sizes
Szabolcs Dávid, Hamed Y. Mesri, Max A. Viergever, Alexander Leemans
This study shows that using ordinary linear least squares (OLLS) in diffusion tensor imaging (DTI) group analyses can inflate effect sizes and increase false positives compared to iterative weighted linear least squares (IWLLS).