Abstract
Jones & Sandland (1974) examined the validity of the linear model for gain per animal on stocking rate. This paper examines the alternative models of Conniffe, Browne & Walsh (1970, 1972) and a proposed extension of the model of Owen & Ridgman (1968). Discrepancies between these models and the linear model are studied in detail.Under specified conditions these models are shown to be very close to the linear model. It is concluded, from the earlier work of Jones & Sandland (1974) and these results, that the empirically justified simple linear model has the advantage of being much easier to fit as well as embodying (approximately) the theoretical advantages of these conceptually superior models under a range of assumptions.