What standard deviation measures
Standard deviation is the most common way to describe how spread out a set of numbers is. A small standard deviation means the values cluster tightly around the mean, while a large one means they are scattered widely. Because it is expressed in the same units as the data, it is easier to interpret than the variance, which is its square. If test scores have a mean of 70 and a standard deviation of 5, most scores sit within a few points of 70. If the standard deviation is 20, scores range far more widely.
Population standard deviation
Use the population formula when your list contains every member of the group you care about, such as the heights of all players on one team or the monthly sales of every store in a small chain. The population standard deviation is written with the Greek letter sigma (σ):
Here μ is the population mean, xi is each value and N is the number of values. You subtract the mean from each value, square each difference so that negatives do not cancel positives, add the squares, divide by N, and take the square root.
Worked example
Take the values 1, 3, 4, 7 and 8. The mean is (1 + 3 + 4 + 7 + 8) / 5 = 4.6. The squared deviations are 12.96, 2.56, 0.36, 5.76 and 11.56, which add up to 33.2. Dividing by 5 gives a variance of 6.64, and the square root gives σ ≈ 2.577.
Sample standard deviation
Most real data is a sample drawn from a larger population: a survey of 500 voters, a batch of parts pulled from a production line, a handful of lab measurements. In that case the sample standard deviation, written s, is the better estimate of the population spread:
The only change is dividing by N − 1 instead of N. This adjustment, called Bessel's correction, compensates for the fact that a sample mean sits closer to its own data than the true population mean does, which would otherwise make the spread look too small. For the five values above, the sample standard deviation is √(33.2 / 4) ≈ 2.881. The difference between the two formulas shrinks as the sample grows.
Standard error and margin of error
The standard error of the mean (SEM) estimates how much the sample mean would vary if you repeated the sampling many times. It equals the standard deviation divided by the square root of the count. Multiplying the SEM by a z multiplier gives a margin of error at a chosen confidence level. The calculator lists eight levels:
| Confidence level | Multiplier |
|---|---|
| 68.3% | 1 × SEM |
| 90% | 1.645 × SEM |
| 95% | 1.960 × SEM |
| 99% | 2.576 × SEM |
| 99.9% | 3.291 × SEM |
| 99.99% | 3.891 × SEM |
| 99.999% | 4.417 × SEM |
| 99.9999% | 4.892 × SEM |
The 95% level is the convention in most fields. The percent of mean column expresses each margin relative to the mean, which is handy for comparing precision across data sets of different scales.
Where standard deviation is used
- Quality control: manufacturers track the standard deviation of part dimensions and flag a process when measurements drift beyond a set number of standard deviations.
- Finance: the standard deviation of returns is a standard measure of an investment's volatility and risk.
- Science and medicine: results are reported as mean ± standard deviation so readers can judge how consistent the measurements were.
- Polling: margins of error in surveys come from the standard error, which is built on the standard deviation.
- Weather and climate: forecasters describe how unusual a temperature or rainfall total is by how many standard deviations it sits from the long-term average.
Reading the frequency table
The frequency table lists each distinct value, how many times it appears and its share of the total. It is a quick way to spot repeated values, the mode, and possible data-entry errors before you rely on the summary statistics.
Things to keep in mind
Standard deviation is sensitive to outliers because deviations are squared: one extreme value can inflate it sharply. For skewed data, consider reporting the median and interquartile range alongside it. Also, the 68-95-99.7 rule (about 68% of values within one standard deviation, 95% within two, 99.7% within three) applies only to data that is roughly normally distributed.