Evaluating permutation-based inference for partial least squares analysis of neuroimaging data.

Finally, while LVs may be significant by permutation testing, it is well established that effect size estimates are crucial for making neurobiological inferences about a finding ( Chen et al., 2017 ), and there are growing concerns that the results from PLS and other multivariate techniques do not necessarily reproduce in held-out samples ( Churchill et al., 2013 ; Dinga et al., 2019 ; Helmer et al., 2024 ; Ji et al., 2021 ; Nakua et al., 2024 ), regardless of significance ( Kovacevic et al., 2013 ; McIntosh, 2022 ).

also in 13,097 other papers