This study was conducted to compare the growth curve models using general linear and multi-level linear growth models and to determine the differences in growth in broiler chickens. For this purpose, a data set containing live weight records of 74 male broiler chickens was used. The measurements were recorded individually, once a week, from hatching to the sixth week. For the analysis of the data, five different growth models, two of the general linear models, and three of the multi-level linear models were used. To find the model that best explains the change; log-likelihood (ll), Akaike information criterion (AIC), Bayes information criterion (BIC), corrected Akaike information criterion (AICC) and Likelihood ratio test (LRT) were used. The results of the study showed that multi-level growth models make more precise predictions than general linear models, and the model that best describes growth is the “random intercept and random slope quadratic growth model” with the smallest fit criteria. According to this model, it was demonstrated that the chickens had significantly different weights since hatching, where the linear and quadratic effect on growth was significant in male broiler chickens, and that individual differences continued significantly during the period of growth.
Ross broiler growth curve hierarchical data repeated measurement covariance structure
Ross etlik piliç büyüme eğrisi hiyerarşik veri tekrarlamalı ölçüm kovaryans yapısı
Birincil Dil | Türkçe |
---|---|
Bölüm | Araştırma Makalesi / Research Article |
Yazarlar | |
Yayımlanma Tarihi | 30 Haziran 2020 |
Yayımlandığı Sayı | Yıl 2020 Cilt: 7 Sayı: 2 |