| 释义 |
Newman–Keuls Statistics.|ˌnjuːmənkœls|
The names of D. Newman (of the Dept. of Statistics, University College, London) and M. Keuls (of the Institute of Horticultural Plant Breeding, Wageningen, Holland), used attrib. to designate a test they devised for assessing the significance of differences between the means of different sets of observations, using the ranges of the sets contributing to the means.
1955D. B. Duncan in Biometrics XI. 26 The Newman–Keuls Test. A test proposed by Newman..in 1939 and again by Keuls..in 1952 succeeds very simply in raising all of the low protection levels of the multiple t test. This test is equivalent to a multiple t test preceded by several preliminary range tests. 1962B. J. Winer Statistical Princ. Exper. Design iii. 85 The level of significance for the Newman–Keuls procedure is considered individually with respect to each test. 1970Jrnl. General Psychol. LXXXII. 159 The Newman Keuls test..was run on the means for the different diagnostic classes.., with results indicating significantly more cartoon enjoyment for normals than for the neurotic and sociopathic Ss. No other differences reached significance. |