How do you address data quality used in mashups? End users can misinterpret results.

A best practice is for IT to publish “authoritative” services as mashables. This provides a level of trust in the data. As for the quality, it depends on the data services.  IT has full control over establishing data quality guidelines for the mashable services.