(f)the consistency between the data sets used for the risk parameters estimation, in particular with regard to the default definition, treatment of defaults, including multiple defaults as referred to in Articles 46(1)(b) and 49, and the sample composition.