Accuracy and Precision

How to Subscribe
MLS & MLT Comprehensive CE Package
Includes 148 CE courses, most popular
$95Add to cart
Pick Your Courses
Up to 8 CE hours
$50Add to cart
Individual course$20Add to cart
Need multiple seats for your university or lab? Get a quote
The page below is a sample from the LabCE course Hematology Instrument Validation/Calibration Verification Protocol. Access the complete course and earn ASCLS P.A.C.E.-approved continuing education credits by subscribing online.

Learn more about Hematology Instrument Validation/Calibration Verification Protocol (online CE course)
Accuracy and Precision

Accuracy and precision are evaluated as part of the validation process. The determination of accuracy and precision are often performed concurrently incorporating the same set of data. The terms accuracy and precision are not synonymous.
Accuracy is the degree of closeness of measurements to the analytes true value. Accuracy is also known as the analytic accuracy (bias). Accuracy determines the systematic error present. It can be verified by using matrix-appropriate reference materials. The materials may be patient samples, altered or unaltered, or other materials with known concentrations.
When evaluating a set of data for accuracy and precision, data reflective of a mean close to the true value of an analyte is an accurate measurement. If the mean of the data set is not close to the true value, the data is less accurate or inaccurate dependent on the distance of the mean from the true value of the analyte.
Precision is a measure of statistical variability. Precision is a description of random error versus the systematic error of accuracy. Precision is the reproducibility and reliability of the test method. It can be verified by repeat analyses of samples containing various concentrations of analytes within a run and between runs over a period of time.
A data set that reflects data points with values that are close to one another is precise. If the data points are scattered and reflective of different values, the data is not precise.
When considering a set of data, the data set can be accurate and not precise, precise and not accurate, both precise and accurate, and neither precise nor accurate. A set of data with a mean close to the to the true value but not close to one another is accurate but not precise. A set of data with values showing reproducibility with the values close to one another but not near the measured value is precise (reproducible) but not accurate.
Two levels of precision may be included in the precision analysis. These are "within run precision" and "between run precision." Within run precision entails multiple analyses of the same sample within one continuous analytic run of samples. Between run precision is multiple analyses of samples over multiple separate runs. The data is analyzed and compared for each type of precision.
Refer to the target diagram for a visual example of accuracy and precision. The bull's eye of the target represents the true value of the test sample.
Calculation of the mean (X), standard deviation (SD), and Coefficient of variation (CV) for determination of accuracy and precision. Instrument manufacturers will provide established values for CV for comparison with analyzed results. Accuracy determination will most often utilize the mean of repeat analyses. Comparison of the mean with the true value will allow for determination of accuracy. Precision is most often determined by the comparison with the achieved CV to the manufacturer's published specifications. Testing results outside of the manufacturer's established limits should be repeated and instrument adjustments as appropriate by manufacturer's recommendations.