The problem of overdispersion as a violation of the assumption of equidispersion in Poisson regression is generally caused by sources of unobserved heterogeneity. missing observations on predictor variables. outliers in the data. errors in the specification of the bridging function. https://www.marketingjeunesse.com/product-category/developer-2-7-vol/