TY - JOUR
AU - David, H. A.
AB -  Abstract Two procedures for ranking variances of normal populations are considered. The first is essentially a gap-test and effects a separation of the variances into two groups whenever the ratio of successive mean squares, arranged in ascending order of magnitude, exceeds a critical tabulated value. The second method is a more elaborate multiple decision procedure based on the distribution of the maximum F-ratio. * I am indebted to B. C. Halliburton of C.S.I.R.O., Prospect, N.S.W., for carrying out the heavy computations underlying Table II, and to Betty Laby of the University of Melbourne for assistance with Table III. My thanks are also due to a referee for a number of valuable comments. 
TI - The Ranking of Variances in Normal Populations*
JF - Journal of the American Statistical Association
DO - 10.1080/01621459.1956.10501353
DA - 1956-12-01
UR - https://www.deepdyve.com/lp/taylor-francis/the-ranking-of-variances-in-normal-populations-94xEC7I0PE
SP - 621
EP - 626
VL - 51
IS - 276
DP - DeepDyve
ER -