How Do You Get Around the Data Mining Fishing Expeditions?

As you have more data points, it seems like the question of interpretation becomes so much more important. How do you get around the sort of data mining fishing expeditions? You can fish and you can eventually find something significant and it may just be a, you know, pure kind of artifact. But I think what a lot of people are trying to do is replication in different data sets because there's so many available now.