LITTLE ROCK, Ark.—When a researcher begins with a conclusion, then looks for data to support that conclusion, “it’s a danger to all of science,” Dr. Paul Hruz, a St. Louis physician-scientist, told a federal judge. Yet he said he has ...