Profile analysis for animal growth data

Abstract

Growth of animal is a repeated measurements data and observations taken on body weight at different time points are correlated to each other. A multivariate technique called profile analysis has been discussed to analyse the growth data (body weight) of animal. This technique is simplified and the data obtained from Jabalpur and Tirupati research stations under AICRP on pigs have been analysed. It was found that the null hypothesis of no interaction of groups with time points and that there is no difference in growth between groups is accepted at 5% level of significance. But the hypothesis of no difference in growth over different time points is rejected for both the research stations. It was observed that the period of 24--32 weeks of age is the period when the growth of pigs is accelerated fast for Jabalpur research station and 12-16 weeks, 20-24 weeks and 28-32 weeks of age are the periods when the growth is accelerated fast for Tirupati research station. These are the periods that need more attention for having the better growth of pigs. Further, 2 traditional (univariate) methods of analysis of repeated measures data have been described and the merits of profile analysis over these methods have also been discussed.