Model Regresi Zero Inflated Poisson Pada Data Overdispersion

Abstract

Overdispersion is a phenomenon of the data variance greater than the average. One of the causes of overdispersion is too many zero value (excess zero) on the response variable. Zero inflated Poisson regression model (ZIP) is one of the method that can be used to overcome problems due to excess zeros. The purpose of this research is to estimate the regression parameters model Zero -inflated Poisson (ZIP) and applying to the data of unsuccessful students in national examinations in senior high school and vocational school in the city of Mataram. Parameter estimation Zero inflated Poisson regression model using the maximum likelihood and maximization expectation algorithm with Newton Rhapson approach. Zero inflated Poisson regression model obtained on the data is:

dan With  is school accreditation; and  is the proportion of teachers who are already certified