ABSTRACT
An experiment comparing 12 methods of feeding chickens (Duckworth and Carpenter; see John and Quenouille, 1977) was done independently in two replicates arranged in different houses. The treatments, forming a 3 × 2 × 2 factorial, were ‘form of protein’, ‘level of protein’, ‘level of fish solubles’. The data are given in Table K.1.
