Thank you for putting this information together and sharing it. I do appreciate the links. One quick comment: I suggest first asking whether your data are "count" data or "metric" (continuous). I view that criterion as the first branch in choosing a distribution you will attempt to fit your data. cheers. zf