{"id":"arxiv/2610.25613","url":"https://arxiv.org/abs/2610.25613","tags":["stat.ME","math.ST"],"score":0,"title":"Anytime-Valid Inference in Linear Models with Clustered Outcomes: Exact Multivariate Tests and Cluster-Robust Confidence Sequences","domain":"stat.ME","source":"arxiv","authors":["Michael Lindon"],"reviews":0,"subject":"stat.ME","summary":"We extend anytime-valid inference for linear models from independent scalar outcomes to units that contribute a vector of dependent outcomes. We develop a parametric and a nonparametric treatment. For the Gaussian multivariate linear model with an unstructured error covariance, we derive a closed-form Bayes factor under the right-Haar prior and a matrix-normal mixture alternative and prove that it is an exact e-value at every sample size and a test martingale in the invariant filtration. To our knowledge, this Bayes factor is new in the generality considered here: a d x q coefficient matrix,…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-10-19","reactions":{}}