© 1968 by Biometrika Trust
Serial factorial design
Rothamsted Experimental Station Harpenden
Serial factorial designs are designs for long-term experiments with one or more basic treatments repeated as separate factors every year. The designs are cyclic, with essentially the same structure in any period of successive years. For example, three rates of a single basic treatment (e.g. nitrogen fertilizer) can be tested each year in a design with just one replicate of the 27 combinations for any three successive years. Methods are described for generating suitable randomized block designs for one or more basic treatments from residue classes of a polynomial ring.