In this study, multiple regression analysis was used to estimate the relationships between egg albumen index and external egg quality traits in Atak-S hens. Egg albumen index was selected as the dependent variable, while egg weight, width, length, shape index, and Haugh unit were selected as independent variables. While the overall model fit was high in the multiple regression analysis, multicollinearity problem were identified among the independent variables. This could affect the model's predictive accuracy, and this problem must be resolved to obtain reliable results. For this purpose, Ridge and Principal Component Regression methods, frequently used in the literature, were applied. In the analyses, the models obtained using the variables used to estimate albumen index and external quality parameters were found to be statistically significant (P<0.05), and the goodness-of-fit coefficient of the models was determined to be R² = 0.88. It was determined that the Ridge regression method yielded slightly more stable results in terms of predictive power, but the Principal Component Regression method provided an advantage in terms of increasing interpretability. These results demonstrate that Ridge and Principal Component Regression methods are robust alternatives that can be used confidently in data structures containing multicollinearity. The findings demonstrate that these methods can be successfully used in poultry breeding and selection studies, yielding more reliable and accurate results. Furthermore, it is clear that these methods will contribute to the development of more effective and robust models in studies on egg quality and productivity.
Gök et al. (Sat,) studied this question.