Evidence factors in observational studies