SIMPLE LINEAR AND MULTIPLE REGRESSION ANALYSES OF MORPHOLOGICAL TRAITS ON BODY WEIGHT IN FEMALE DORPER SHEEP LAMBS

Abstract

Regression is a very influential statistical technique that uses independent variables to describe the dependent variable. The present study was performed to determine the best-fitted model from different regression techniques to predict body weight (BW) from morphological traits viz. heart girth (HG), rump height (RH), body length (BL), withers height (WH) and sternum height (SH). Twenty-eight female Dorper sheep lambs at birth were used for data collection. The simple and multiple regression were used for data analysis. The coefficients of determination (R2) and mean square error (MSE) were used to determine the best-fitted regression model. The results indicated that in simple regression the best fitted regression model for estimation of BW in female Dorper lamb was the model including BL (R2 = 0.79, MSE = 1.43), in multiple regression, the best fitted regression model for estimation of body weight in female Dorper lamb was model including HG, RH, BL, WH, SH (R2 = 0.89, MSE = 0.90) in the study. The results of simple regression suggest that BL can truly estimate the body weight in female Dorper sheep lambs. Multiple regression findings suggest that two or more morphological traits can truly estimate body weight in the female Dorper lambs. The study will help the farmers to accurately predict the body weight of the Dorper sheep lambs using the morphological traits.