Stepwise Method and Factor Scoring in Multiple Regression Analysis of Cashmere Production in Liaoning Cashmere Goats

Abstract

Liaoning cashmere goat is a well-known local cashmere goat breed in China and even in the world. It is famous for producing cashmere with superior quality and high yield. Cashmere yield, body measurements, and body weight are the primary indicators of cashmere goat breeding, but the correlation between them is not yet clear. Therefore, this study investigated the relationship between certain body measurements, body weight, and cashmere yield in Liaoning cashmere goats using stepwise and factor score analyses in a multiple regression analysis. For this purpose, the body measurements (body slanting length (BSL), body height (BH), chest circumference (CC), pipe circumference (PC), chest depth (CD), chest width (CW), hip breadth (HB), body weight (BW) and cashmere yield (CY)) of 200 (2-year-old) Liaoning cashmere goats were collected. Stepwise analysis of the results showed that body weight had the greatest direct effect on cashmere yield, followed by hip breadth, while chest circumference mainly affected cashmere yield indirectly. The results of factor score analysis showed that the independent variable can be represented by two factors, which explained 49.596% and 12.095% of the total variance, respectively. The factor scores used in the regression analysis explained 75.8% of the total variance in Liaoning cashmere yield. The above studies show that the growth traits of Liaoning cashmere goats are closely related to the cashmere yield. Growth traits should be considered important factors in breed selection, germplasm identification, and rearing.

A new strategy for the treatment of hepatitis B patients based on data mining

“…The stepwise regression is performed using the principal components of gene expression in the pathway and the related receptors. The combination of forward variable selection method and backward variable elimination method makes it possible to find eligible independent variables without introducing independent variables that have no significant influence on the dependent variables 27 .…”