About the Coefficient of Variation Calculator
This coefficient of variation calculator measures how spread out a data set is relative to its mean. Paste your numbers and it returns the coefficient of variation (CV) — the standard deviation divided by the mean, expressed as a percentage — along with the mean, standard deviation and variance. CV is also called the relative standard deviation (RSD).
Because CV has no units, it lets you compare variability between data sets measured on different scales or with very different averages: the consistency of two production lines, the precision of lab assays, the volatility of investments relative to their return, or the stability of monthly sales across stores of different sizes.
Choose sample if your numbers are a sample from a larger population (n − 1 in the standard deviation, the usual choice) or population if they are the complete set. CV is only meaningful for ratio-scale data with a positive mean; it is undefined when the mean is zero.
With the default inputs, the coefficient of variation is 2.2175%. Change any value above to recalculate instantly.
How to use the coefficient of variation calculator
- 1Paste or type your data values.
- 2Choose sample (most common) or population.
- 3Read the coefficient of variation as a percentage.
- 4Compare CVs between data sets to see which is relatively more variable.
Formula and method
The coefficient of variation divides the standard deviation by the mean, turning an absolute spread into a relative one. A CV of 5% means values typically differ from the average by about 5% of the average, regardless of the units.
The sample standard deviation s = √[Σ(x − x̄)² ÷ (n − 1)] is used for samples; the population version divides by n instead. If the mean is negative the absolute value of the mean is used. CV should only be used for data measured on a ratio scale with a true zero (lengths, weights, prices), not for temperatures in °C or other interval scales.
- CV
- Coefficient of variation (relative standard deviation)
- s
- Standard deviation (sample or population)
- x̄
- Arithmetic mean of the data
Worked examples
Precision of eight lab measurements
The eight readings average 12.0375 with a sample standard deviation of about 0.2669, so CV = 0.2669 ÷ 12.0375 ≈ 2.22%. That is a tight, highly repeatable measurement.
Monthly sales for a small store (population)
Treating the six months as the whole population, the mean is about 4,616.67 and the standard deviation about 794.60, giving CV ≈ 17.2% — moderate month-to-month variability.
Comparing volatile returns
These returns average 7% with a sample standard deviation of about 7.43 percentage points, so the CV is about 106.1% — the variability is larger than the average itself, a sign of high relative risk. Because returns can be negative and the mean can approach zero, treat CV on returns as a rough risk-per-unit-of-return gauge rather than a precise measure.
Frequently asked questions
What is a good coefficient of variation?+
It depends on the field. In lab assays a CV under 5–10% is often considered good precision, while in finance or biology much higher CVs are normal. As a rough guide, under 10% is low variability, 10–30% moderate and over 30% high.
Is coefficient of variation the same as relative standard deviation?+
Yes. Relative standard deviation (RSD, or %RSD) is the coefficient of variation expressed as a percentage. Both are the standard deviation divided by the absolute mean.
When should I not use the coefficient of variation?+
Avoid it when the mean is zero or close to zero (CV explodes), be cautious when data can be negative (the mean may sit near zero), or for interval scales without a true zero such as Celsius temperatures, where the ratio to the mean has no meaning.
Should I use sample or population standard deviation?+
Use the sample version (dividing by n − 1) when your data are a sample drawn from a larger group, which is the usual case. Use the population version only when the data include every member of the group you care about.
Why use CV instead of standard deviation?+
Standard deviation is in the units of the data, so it cannot fairly compare, say, the variability of house prices with that of rents. CV is unitless and scaled by the mean, so it compares relative consistency across different scales.