基于二元Logistic回归模型的德国牧羊犬幼犬自然行为表现与胆量相关性分析OA

To enhance the objectivity and accuracy of selecting working dog puppies,this study employed a binary Logistic regression model to investigate the relationship between five behavioral factors and boldness in 2-month-old German Shepherd puppies under natural condi-tions.Videos of natural behaviors were collected from 60 German Shepherd puppies in a closed and undisturbed environment.Then,the fre-quency of exploratory behavior,reaction to unfamiliar objects,barking frequency,immobility duration,and frequency of social interaction were selected as independent variables.Finally,the boldness of the puppies was assessed by an evaluation team,and the resulting scores served as the dependent variable in the analysis.The results showed as follows:The frequency of exploratory behavior and reaction to unfamil-iar objects were significant predictors of boldness assessment(P＜0.05).Puppies with higher frequencies of exploratory behavior and calmer reactions to unfamiliar objects achieved higher scores in the boldness test.Priority feeding behavior also significantly predicted the boldness of the puppies(P＜0.05);and the puppies that fed first were 8.045 times more likely to be rated as being bold,compared with those that fed later.This study offered significant guidance for screening candidate police dogs,might serve as an objective basis for the early selection of bold German Shepherd puppies for working purposes,and contributed to optimizing the breeding process and improving selection quality.