方法对比

	Fisher精确随机推断	Bootstrap Inference	Jackknife Resampling	分位数回归（非参数变体）
领域	统计学	统计学	统计学	统计学
方法族	Regression model	Regression model	Regression model	Regression model
起源年份	1935	1979	1956	1978
提出者	Ronald A. Fisher	Bradley Efron	Quenouille (1956); reviewed by Miller (1974)	Koenker & Bassett
类型	Exact permutation-based inference	Resampling-based inference	Resampling / bias and variance estimation	Quantile regression (nonparametric variants)
开创性文献	Fisher, R. A. (1935). The Design of Experiments. Oliver & Boyd.	Efron, B. (1979). Bootstrap Methods: Another Look at the Jackknife. Annals of Statistics, 7(1), 1-26.	Quenouille, M. H. (1956). Notes on Bias in Estimation. Biometrika, 43(3/4), 353-360.	Koenker, R. & Bassett, G. (1978). Regression Quantiles. Econometrica, 46(1), 33-50.
别名	fisher randomization test, permutation inference, exact randomization test, randomizasyon çıkarımı (fisher exact randomization)	bootstrap, bootstrap resampling, nonparametric bootstrap, Bootstrap Çıkarımı	leave-one-out resampling, Quenouille-Tukey jackknife, delete-one jackknife, Jackknife Yeniden Örnekleme	quantile regression, median regression, distribution-free quantile regression, Kantil Regresyon (Nonparametric Varyantlar)
相关	5	5	5	5
摘要	Randomization inference, introduced by Ronald A. Fisher in The Design of Experiments (1935), computes an exact p-value by evaluating a test statistic across all possible treatment assignments under Fisher's sharp null hypothesis. It is regarded as the gold standard for analysing designed experiments because its validity rests on the known assignment mechanism rather than on distributional assumptions.	Bootstrap inference, introduced by Bradley Efron in 1979, estimates the sampling distribution of a statistic by repeatedly resampling the observed data with replacement. It requires no distributional assumption and produces reliable confidence intervals even in small samples.	The jackknife is a classical resampling method that estimates the bias and variance of a statistic by systematically recomputing it with one observation left out at a time. Introduced by Quenouille in 1956 and later reviewed by Miller in 1974, it predates the bootstrap and remains a simple, deterministic tool for assessing estimator stability.	Quantile regression, introduced by Koenker and Bassett in 1978, models a chosen conditional quantile (such as the median or the 25th and 75th percentiles) of a continuous outcome rather than its mean. Its nonparametric variants fit these quantile relationships without assuming a distribution for the errors, making them a robust complement to mean-based regression on skewed data.