Beta four parameter GLMM approach to evaluate paddy productivity

Abstract

Abstract Indonesia is one of the top five counties in the world that produces rice. In average, Indonesia has a paddy productivity of 5.2 tons per hectare and continuously growing within 1.59% annual rate. Eventhough high, one of the main problems faced is that Indonesia is a vast archipelago country that has a diverse agro-climatic and socio-economic conditions which can significantly impact paddy productivity in an area. Given also that the paddy productivity in Indonesia has a certain minimum and maximum value, thus using a beta four parameter-based model is considered appropriate. We have proposed novel Beta Four Parameter GLMM that was an extension of the Beta Four Parameter regression model developed by Zou and Huang (2022). The model was applied in an empirical study based on farmers crop cutting survey in Central Kalimantan Indonesia. Results show that the proposed Beta Four Parameter GLMM has a better model fit. The model showed that pest attacks, pest handling, and resource allocation was considered the most influential variable affecting paddy productivity. Hence, it is suggested that the government must focus to construct a comprehensive strategy to improve paddy productivity by making it more resilient to pest and having more sufficient resource allocation management.