Residual randomization is a method for testing and inference in regression models based on invariance assumptions on the errors; e.g., inference assuming only exchangeability of errors, or sign symmetry, or both. Compared to standard normal OLS these assumptions are weaker, which adds robustness. Compared to bootstrap, residual randomization is more flexible, it does not rely on asymptotic normality, and addresses the inference problem in a unified way.
Papers
Life after bootstrap: residual randomization inference in regression models (Working paper, 2019, pdf, slides)
Introduction to Residual Randomization: The R Package RRI (Technical report, 2019, pdf)
Code
RRI R package: CRAN https://cran.r-project.org/web/packages/RRI/index.html