Visual Tools
Calculators
Tables
Mathematical Keyboard
Converters
Other Tools

How to Find the Standard Deviation of a Random Variable

The standard deviation is the square root of the variance. It measures typical distance from the mean in the same units as the variable itself, which makes it the number you report.

Recognize the task

The question asks for σ\sigma, "typical deviation", spread in the original units, or needs a value to standardize with (z=x−μσz = \frac{x - \mu}{\sigma}). Typical inputs: a pmf table, a variance already found, a named distribution such as a binomial count. Test: do you have, or can you get, the variance? Then the standard deviation is one square root away.
The relation between the two spreads is explained in the lesson section Variance vs Standard Deviation.

Two ways to do it

Square root of the varianceusual
σ=Var⁡(X)\sigma = \sqrt{\operatorname{Var}(X)}
Find the variance first, then take the square root.
Formula of a named distributionshortcut
X∼Bin(n,p)  ⇒  σ=np(1−p)X \sim \text{Bin}(n, p) \;\Rightarrow\; \sigma = \sqrt{np(1 - p)}
Take the square root of the distribution variance.

Steps

StepOn X∈{0,1,2,3}X \in \{0, 1, 2, 3\} with probabilities 0.1, 0.3, 0.4, 0.20.1,\ 0.3,\ 0.4,\ 0.2
1Find the variance. Mean, mean of the square, subtract.Var⁡(X)=3.7−1.72=0.81\operatorname{Var}(X) = 3.7 - 1.7^2 = 0.81
2Take the positive square root. Round only at the end; keep exact fractions or radicals as long as you can.σ=0.81=0.9\sigma = \sqrt{0.81} = 0.9
3Attach the units and compare with the range. The standard deviation has the units of XX and is at most half the range of the values.σ=0.9\sigma = 0.9, less than half of 3−03 - 0 ✓

Worked examples

Example 1 · a fair die

exact first
Var⁡(X)=3512\operatorname{Var}(X) = \frac{35}{12}
from the variance page
σ=3512\sigma = \sqrt{\frac{35}{12}}
square root
σ≈1.7078\sigma \approx 1.7078
rounded to 4 places

Example 2 · heads in 100 fair coin flips

binomial
X∼Bin(100,0.5)X \sim \text{Bin}(100, 0.5)
fixed trials, fixed pp
Var⁡(X)=100⋅0.5⋅0.5=25\operatorname{Var}(X) = 100 \cdot 0.5 \cdot 0.5 = 25
np(1−p)np(1 - p)
σ=5\sigma = 5 heads
square root

Example 3 · the sum of two independent variables

add variances
σX=3, σY=4\sigma_X = 3,\ \sigma_Y = 4, independent
Var⁡(X+Y)=9+16=25\operatorname{Var}(X + Y) = 9 + 16 = 25
variances add, not deviations
σX+Y=5\sigma_{X+Y} = 5
square root of the sum
For counts from a named distribution, see binomial probability.

See it happen

The variance is in squared units (dollars squared, minutes squared), which is hard to picture. The square root brings the spread back to the scale of the variable, so you can mark μ±σ\mu \pm \sigma on the same axis as the values. For bell-shaped distributions most of the probability sits within two standard deviations of the mean. Compare spreads in the variance visualizer.

Common mistakes

Adding standard deviations

✗ σX+Y=3+4=7\sigma_{X+Y} = 3 + 4 = 7
✓ σX+Y=9+16=5\sigma_{X+Y} = \sqrt{9 + 16} = 5 for independent X,YX, Y
Only variances add (variance of a sum).

A negative standard deviation

✗ σ−3X+1=−3⋅2=−6\sigma_{-3X + 1} = -3 \cdot 2 = -6 when σX=2\sigma_X = 2
✓ σ−3X+1=∣−3∣⋅2=6\sigma_{-3X + 1} = |{-3}| \cdot 2 = 6

Averaging absolute distances

✗ σ=E ∣X−1.7∣=0.76\sigma = E\,|X - 1.7| = 0.76
✓ σ=E[(X−1.7)2]=0.9\sigma = \sqrt{E[(X - 1.7)^2]} = 0.9
The mean absolute deviation is a different, smaller measure.