Pao-lu Hsu or Baolu Xu, Chinese statistician (Beijing 01 September 1910 – Beijing 18 December 1970)

Proposed Hsu method

Contribution to the theory of Student’s t-test as applied to the problem of two-samples. Statis. Res. Mem.2:1-24, 1938

Gave important results on the optimal quadratic estimates of variance and variance components

On the best unbiased quadratic estimate of the variance. Statis. Res. Mem. 2:91-104, 1938

Established the first optimal hypotheses in the likelihood-ratio test for univariate linear models or the first nonlocal optimal property for any hypotheses specifying the value of more than one parameter

Analysis of variance from the power function standpoint. Biometrika 32(1):62-9, 1941

A general weak limit theorem for independent distributions (1947)

Hsu-Robbins-Erdos theorem

With H. Robbins. Complete convergence and the law of large numbers. Proceed. Natl. Acad. Sciences USA 33(2):25, 1947

Obtained several theorems in the matrix theory

Obtained necessary and sufficient conditions for the Markov estimator to be the best unbiased second-order estimator

Eponym of P.L. Hsu theorem (about optimal property for F-test) and Hsu’s method or Δ method