This actually matches a dataset we are looking to publish (soon) very well - both with regards to frequency of inappropriate duplications and the breakdown of types. I think we may have to get used to the idea that fraudulent studies are much more common than previously thought.