"what DO scientists do with recalcitrant data?"

If by recalcitrant, you mean data that does not support an existing hypothesis, then they back up and revise the hypothesis. Nearly all great advances in science happen because of unexpected (relcalcitrant?) data.

Remember the story of Galileo dropping two balls of different weights off the tower? He fully expected them to fall at different speeds. Instead, I believe he got recalcitrant data.