Abstract
Influential observations influence the Poisson regression model (PRM) inferences. There are the situations in the PRM, where the explanatory variables are correlated and influential observations occurs simultaneously. So the Poisson ridge regression model (PRRM) is proposed to reduce the effect of multicollinearity. This study proposes some influence diagnostics for the PRRM to identify the influential observations. The performance of proposed PRRM diagnostic methods is evaluated through Monte Carlo simulation study and two real applications. The simulation and real applications results show the superiority of proposed diagnostic methods over maximum likelihood estimation based diagnostic methods.