Protein quality of feeding-stuffs

Abstract

1. The reproducibility of an assay for methionine based on live-weight gains in young chicks has been tested in a collaborative study in six laboratories.2. Results were analysed by the slope-ratio procedure and, in general, the assays were statistically valid. The variability between laboratories was similar to that found in previous studies of variablity within a single assay in one laboratory.3. With the combined estimates from five or six laboratories the standard error of the estimates was apporoximately 10% of the mean. Expressing response as‘g gain/g food eaten’ gave no more precision than using ‘weight gain’ alone, but is nevertheless thought to be less open to error due to apetite effects.4. The experiments have shown that materials can be ranked consistently, but that the absolute estimates of potency varied between assays and further improvements are desirable if potency estimates are to be used for the calibration of in vitro procedures of protein evaluation.