if you were going to use your standard curve in a clinical setting, such as drug testing, do you think your percentage error is reasonable?