Data which constitute satisfactory evidence in other contexts are sometimes not treated as valid evidence in the context of scientific confirmation. I give a justificatory explanation of this fact, appealing to the incentives, biases, and social situatedness of scientists.