References; Linked-In Posts; White Papers
How to measure Uncertainty of Measurement
When you start working with all the numbers, you see uncertainties, or confidence limits, everywhere. Our goal is to determine if the patient or QC mean and SD changed significantly. which makes uncertainty of measurement critical ... as are significant figures, by the way.
If we measure the percent low and high risk patients each day, week and month for HbA1c, there are uncertainties in each data set. For instance, in EP Evaluator software, they calculate confidence limits for the mean and SD saying: "95% Confidence Interval for SD: In a sense, this measures the "precision of the precision". It shows how much the SD might vary if the precision experiment was repeated with different experimental results. The width of the confidence interval depends on two factors: N and the intrinsic SD of the system. The reliability of the precision estimate improves as N increases."
If you have other comments or questions about MU, please send them. Or please join us on Thursday at 2 PM for the online meeting.