You have a dataset with a known error. How do you quickly calculate the impact of that error on the final average score?