Calculate control means, SD and CV
Summarise a series of control measurements and calculate statistical control limits. Use a stable, representative data set and your laboratory’s approved review process.
Last reviewed
How it works
What the calculation tells you
The mean describes the centre of the entered measurements. The sample standard deviation, or SD, describes their spread in the same units as the results. The coefficient of variation, or CV, expresses that spread relative to the absolute mean as a percentage. These quantities answer different questions: a stable mean does not prove low imprecision, and a small CV does not prove that results are close to a reference value.
Prepare a suitable series
Use results from the same control material, concentration level and measurement procedure. Check the dates, reagent conditions, calibration history and units before combining observations. A collection that includes an unresolved shift or a change of control lot may not describe one stable process. Review unusual measurements against documented events rather than removing them simply because they widen the calculated limits. Keep the original results and the reason for any exclusion in your local records.
Enter measurements separated by spaces, semicolons, commas or line breaks. Use a decimal point for fractional values: a comma acts as a separator, not as a reliable decimal marker. The parser ignores entries that do not convert to finite numbers, so compare the displayed count with the number you intended to enter. At least two valid measurements are needed. Two observations allow arithmetic, but are not enough to establish dependable long-term control limits.
Formulas and rounding
The mean is the sum of the measurements divided by n. The sample variance is the sum of squared deviations from that mean divided by n − 1; SD is its square root. Using n − 1 distinguishes this calculation from a population SD. CV is 100 × SD / |mean|. When the mean is zero, CV is undefined, and when it is near zero, CV can be misleading even if the arithmetic is possible.
Limits are calculated as mean ± one, two and three SD. Keep full precision during calculation and round only the displayed results. If you reconstruct limits from an already rounded SD, small differences from the displayed limits are expected. The bands describe statistical distance; they are not allowable total error limits, patient reference intervals or a substitute for a validated QC strategy.
Use the results in context
Inspect the sequence as well as the summary. A gradual drift and a sudden shift can produce similar overall statistics but require different investigations. Confirm that the estimate represents routine operating conditions and the concentration range of interest. A short teaching series is useful for checking arithmetic, not for approving a new control lot. Establish or revise assigned statistics through the laboratory’s own documented procedure, including review, authorisation and subsequent monitoring. This calculator supports that review; it does not decide whether patient results may be released.
Worked example
Values: 5.1, 4.9, 5.0, 5.2, 4.8, 5.0, 5.1, 4.9, 5.0, 5.0.
The sum is 50.0, so n = 10 and mean = 5.00. Squared deviations sum to 0.12; sample SD = √(0.12 / 9) = 0.115 after rounding. CV = 2.31 %. Using unrounded values, ±2 SD limits are 4.77–5.23 and ±3 SD limits are 4.65–5.35.