Usability of the Factor Analysis Scores in Multiple Linear Regression Analyses for the Prediction of Daily Milk Yield in Norduz Goats

Abstract

The aim of this study was to determine the relationship between daily milk yield and udder traits using multiple regression analyses in order to predict daily milk yield in Norduz goats. 10 udder traits including upper udder height, bottom udder height, udder depth, udder width, udder circumference, left teat length, right teat length, left teat circumference, right teat circumference and teat angle. The data was collected from 27 Norduz goats raised in pastoral conditions in the Norduz region of Van province South Eastern Turkey. Factor analysis was employed to simplify the complex relationships between udder traits. After the udder traits were exposed to factor analysis, four factors with Eigen values greater than 1 were selected as explanatory (independent) variables and used for multiple linear regression analysis. First factor was named teat factor, second and third factors were named udder factors while the fourth was udder bottom height. The 2nd and 3rd factors, which were significant, were then used to fit the regression model. The study found that two udder factors had significant statistical effect on daily milk yield and these factors together had accounted for 78.6 % of the variation in daily milk yield. The findings of this study showed that both multivariate and univariate approaches can be used to determine the relationship between milk yield and udder traits. In addition, these statistical approaches may also be useful to eliminate multicollinearity problems among large number of variables. In conclusion, the study proved that both univariate and multivariate methods can be applied successfully to predict daily milk yield using udder traits in goats.