Latin rectangle designs for 2n factorial experiments on 32 plots
- 1 October 1951
- journal article
- research article
- Published by Cambridge University Press (CUP) in The Journal of Agricultural Science
- Vol. 41 (4) , 315-316
- https://doi.org/10.1017/s0021859600049492
Abstract
Designs are given for factorial experiments with three, four or five factors at two levels each, using thirty-two plots in a 4 × 8 lay-out on the ground. The effects of both rows and columns can be eliminated from the estimate of error. Provided that three-factor interactions can be ignored, information can be retained on all two-factor interactions.Keywords
This publication has 0 references indexed in Scilit: