Abstract
The information on treatment contrasts, provided by an incomplete block experiment, is given by a method that involves the inversion of a matrix of the form A + aHH′ where a + 0. It is shown that, for many of the standard designs, the contracts can be easily obtained by an appropriate choice of a. Intrablock estimates and the recovery of interblock information are considered.

This publication has 0 references indexed in Scilit: