Abstract
The chain block design is an incomplete block design in which large numbers of treatments can be compared with few replications. The author extends these designs in such a way that, in addition to corrections for block differences, effects of ordering within blocks may be estimated and removed. This is called two-way elimination. Several design patterns are given and the calculations are illustrated with an expt. in testing of tires for rate of treadwear.

This publication has 1 reference indexed in Scilit: