Eliminating accidental deviations to minimize generalization error and maximize replicability: Applications in connectomics and genomics

Reproducibility, the ability to replicate scientific findings, is a prerequisite for scientific discovery and clinical utility. Troublingly, we are in the midst of a reproducibility crisis. A key to reproducibility is that multiple measurements of the same item (e.g., experimental sample or clinical participant) under fixed experimental constraints are relatively similar to one another. We demonstrate that existing reproducibility statistics, such as intra-class correlation coefficient and fingerprinting, are not valid measures of reproducibility, in that they can provide unreasonably low or high results, even without model misspecification. We therefore propose a novel statistic, discriminability, which quantifies the degree to which an individual’s samples are relatively similar to one another, without restricting the data to be univariate, Gaussian, or even Euclidean. Using this statistic, we introduce the possibility of optimizing experimental design via increasing discriminability and prove that optimizing discriminability improves performance bounds in subsequent inference tasks. In extensive simulated and real datasets (focusing on brain imaging and demonstrating on genomics), only optimizing data discriminability improves performance on all subsequent inference tasks for each dataset. We therefore suggest that designing experiments and analyses to optimize discriminability may be a crucial step in solving the reproducibility crisis, and more generally, mitigating accidental measurement error.