Determination of best-fitted regression model for prediction of body weight in attappady black goats

T.V. Raja, R. T. Venkatachalap

Abstract

The present investigation was undertaken to determine the best-fitted regression model for the prediction of body weight using different linear body measurements of Attappady Black goats in Kerala, India. Data on body weight and body measurements recorded on 1412 Attappady Black goats consisting of 493 males and 919 females from its breeding tract were used for the study. The whole data was classified into five age groups viz., 0-3, >3-6, >6-12, >12 and 0-12 months. The simple and multiple regression models were fitted with body weight as dependant variable and chest girth, body length and height at withers as independent variables. The correlation coefficients between body weights and body measurements at different ages were positive and strongly correlated (P<0.01) ranging from 0.509 to 0.915. The chest girth alone accounted for a maximum of 79.4 and 77.9 per cent of the total variation in body weight in males and females, respectively, signifying its importance for the prediction of body weight. Highest coefficient of determination (R2 - value) was obtained in the models when all the body measurements were included. These models also showed smaller RMSE and standard deviation ratio (SDR) thereby registering a better goodness of fit.