Philippine E-Journals

Application of Factor Analysis Scores in a Multiple Linear Regression Model for the Prediction of Liveweight in Immature West African Dwarf Goat

Ogah Danlami Moses

Discipline: Veterinary Medicine, Animal Science

 

Abstract:

The study was conducted to find out the reliability of factor analysis scores in multiple linear regression model that were used to estimate body weight using various body measurements – wither height, rump height, body length, foreleg length, shoulder length, rump length, hearth girth, canon circumference and chest depth obtained from 125 approximately one-year-old West African dwarf goats (45 males and 80 females). There was significant sex effect on all the traits in favour of males except for body length which favoured female goats. The factor analysis result showed 3 factors with eigen values greater than 1 for both males and females and were selected as explanatory variables and used to estimate body weight of young goats in multiple linear regression model. Result of multiple linear regression using body measurements for both males and females exhibits multicollinearity of the variables with high variance inflation factor (VIF) necessitating use of factor scores. Two of the three factors scores were significant (P<0.001) (P<0.05) in the estimation of male and female body weights, respectively. The factors accounted for 86.6% and 75.2% of the total variation in the body weight in males and females, respectively.