Multicollinearity Problem and Bias Estimates in Japanese Quail
Multicollinearity Problem and Bias Estimates in Japanese Quail
Mustafa Sahin1,*, Esra Yavuz1 and Fatih Uckardes2
ABSTRACT
The aim of this study is to estimate the albumen index value that is one of the interior egg quality traits using the Least Squares (LS), Ridge Regression (RR) and Principal Components Regression (PCR) methods. For this reason, eggs were collected from 20 to 24 weeks of age, non-selected and random-mated Japanese quails. In this study, we investigated the characteristics of egg external quality such as egg width (X1), egg length (X2), Haugh unit (X3), and shape index (X4). The impact of these variables on albumen index was analyzed. Due to the multicollinearity between independent variables, the standard errors and VIF values of the partial regression coefficients in the LS method were found to be quite high. Therefore, the RR and PCR methods were used as an alternative to LS. Compared with LS, PCR and RR led to small standard deviations and more accurate and more reliable partial regression coefficients. As a result, it was concluded that the use of RR and PCR analysis methods could be more accurate instead of the LS method under multicollinearity problem.
To share on other social networks, click on any share button. What are these?
April 2018
Vol. 50, Iss. 2, Pages 401-797
