Application of Generalized Estimating Equations (GEE) and Random effect model for estimation of occurrence and non occurrence of mastitis In cattle and buffaloes
115 / 138
Keywords:
Mastitis, Binary, generalized estimating equations, Random effect modelAbstract
In this paper we have studied the effect of age, season of calving and species of animals on the occurrence and non occurrence of mastitis by using the longitudinal data. The data on the month wise occurrence /non occurrence of mastitis recorded for the period 2005 to 2014 from the dairy farm of Indian Veterinary Research Institute (IVRI), Izatnagar,Bareilly, Uttar Pradesh. The data on 225 dairy animals have been used in the present study. To analyze the effect of different factors on the mastitis occurrence we have used generalized estimating equations (GEE) and random effects model. Our study revealed that the effect of age on occurrence of mastitis was significant. However there was no effect of seasons (summer, winter and rainy) and species on the mastitis occurrence. The interactions of age, seasons and species with time were non-significant indicating that the factors do not influence the mastitis occurrence over the time. The generalized estimating equations model proved to be superior to the random effect model as the error is less for every parameter in the GEE model.Â