Can you describe a time when an experiment you were running consistently failed, and how you went about troubleshooting it for Duke University?
