A Method of Increasing Precision in Vitamin A Assay

Abstract

In a study of methods for the improvement of the bio-assay for vitamin A, the average standard deviation from the mean gain has been chosen as the criterion in making comparisons. The possibility of improving the bio-assay for vitamin A by strict laboratory standardization of the generally accepted method has been demonstrated. Further reduction of the variability of the assay has been effected through two channels. First, animals were eliminated from the assay group that gave indications in the depletion period of erratic response to the feeding of the test substance. These eliminations were made on the basis of estimations of future gain by means of a regression equation calculated on certain data collected in the depletion period and later gain in the test period. Second, the uniformity and thence the reliability of the assay has been increased by shortening the test period from 8 to 5 weeks. The adoption of the two procedures in conducting the assay increased the information yielded by the test 127%.