{"id":"arxiv/2611.23775","url":"https://arxiv.org/abs/2611.23775","tags":["stat.ME","math.ST"],"score":0,"title":"Anytime-valid goodness-of-fit test by $e$-spacings","domain":"stat.ME","source":"arxiv","authors":["Yunda Hao","Kin Wai Chan"],"reviews":0,"subject":"stat.ME","summary":"To perform an anytime-valid goodness-of-fit test, we propose spacing-based \\(e\\)-processes that control the type-I error rate under arbitrary stopping rules. Based on the maximum spacing of order statistics, we construct spacing-based \\(e\\)-processes for nonparametrically testing distribution. The resulting procedures are simple and computationally efficient, and are power-consistent. We call them $e$-spacing tests. This approach is especially powerful in detecting localized discrepancy between the data-generating density and the null density. Under mild conditions, the logarithm of the…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-11-18","reactions":{}}