The main objective of this paper was to illustrate application of Heckman-Conway-Maxwell-Poisson ‘HeckCOMPoisson’ model for analyzing corruption. HeckCOMPoisson model is an integrated model from Heckman and Conway-Maxwell-Poisson model. HeckCOMPoisson was developed to handle count model with selection. The specific objective of this paper was to test the fitness of the integrated HeckCOMPoisson model in predicting number of people paying bribes. Corruption is among the most serious problems of the societies because it destroys the democratic progress by creating unemployment, increasing poverty level among other negative effects. To achieve the objective of this paper, summary tables were created, graphs were drawn and the results discussed extensively. This paper used the corruption data of 2019 from Ethics and Anti-Corruption Commission (EACC) of Kenya. The results of this study showed that HeckCOMPoisson perform perfectly in terms of Goodness-of-Fit (GOF) and in prediction of count data with selection and handling under-dispersion and over-dispersion experienced in count data. The results of this study proved that HeckCOMPoisson distribution is flexible and gives robust models for dispersed counts. By parametrizing HeckCOMPoisson distributions via the mean, variance and model prediction, this study placed HeckCOMPoisson models on the same level of interpretability and parsimony as other popular, competing count models.
Leonard King'ora Thuo , Dr. Joseph Ouno Omondi , Dr. Boniface Kwach "Application of Heckman - Conway - Maxwell - Poisson Model for Analysing Corruption" Iconic Research And Engineering Journals Volume 6 Issue 2 2022 Page 101-112
Leonard King'ora Thuo , Dr. Joseph Ouno Omondi , Dr. Boniface Kwach "Application of Heckman - Conway - Maxwell - Poisson Model for Analysing Corruption" Iconic Research And Engineering Journals, 6(2)