If presented with conflicting data in an experiment, how would you approach resolving the discrepancies?
