Use of Factor Scores in Multiple Regression Model for Predicting the Live Weight of Native Chickens using Body Measurements

In this study, factor and multiple regression analyses were combined to estimate Live weight from ten body measurements of the Nigerian native chickens. A total of 1500 (750 males and 750 females) were used in this investigation. Sexual dimorphism existed between the sexes. Kaiser-Meyer-Olkin (KMO) and Bartlett’s test of sphericity were used to test the appropriateness of factor analysis on the data and they found fit for application in factor analysis. In the Varimax rotation factor analysis, three factors with eigenvalues greater than 1 were extracted which accounted for 49.63% of total variation for the males, while five factors with eigenvalues greater than 1 which accounted for 62.01% of total variation were extracted for the female data set. When utilized as independent variables in multiple regression analyses, two factors each, had positive significant effects on Live weight, accounting for 86.4% and 75.3% variation in Live weight for males and females, respectively. The use of some Body measurements might provide useful information on improvement of Live weight in chicken breeding programme. Keywords: Native chicken, Factor analysis, Varimax rotation, regression, Multicollinearity