How standard deviation measures data spread
Standard deviation quantifies how far individual values typically stray from the average. It's the most common measure of variability in statistics, used across science, finance, quality control, and research. A small standard deviation means data clusters tightly; a large one signals scattered, diverse values.
Variance—the square of standard deviation—serves a similar purpose but in squared units, making it less intuitive for everyday interpretation yet mathematically convenient.
The formula
σ = √[ Σ(x − μ)² / N ] for population; s = √[ Σ(x − x̄)² / (n − 1) ] for sample
For population: sum the squared differences from the mean, divide by total count N, then take the square root. For sample: divide by n−1 instead of n (Bessel's correction).
Worked example
Imagine a teacher records test scores for a class: 72, 85, 78, 92, 88.
Step 1: Calculate the mean
Mean = (72 + 85 + 78 + 92 + 88) ÷ 5 = 415 ÷ 5 = 83
Step 2: Find squared differences from the mean
- (72 − 83)² = 121
- (85 − 83)² = 4
- (78 − 83)² = 25
- (92 − 83)² = 81
- (88 − 83)² = 25
Sum = 121 + 4 + 25 + 81 + 25 = 256
Step 3: Divide by n (or n−1 for sample)
Assuming this is a sample (representing a larger population of students):
256 ÷ (5 − 1) = 256 ÷ 4 = 64
Step 4: Take the square root
√64 = 8
Result: Sample standard deviation = 8 points
Variance = 64 square points
This means scores typically deviate about 8 points from the mean of 83. Most students' scores fall between 75 and 91.
If this were the entire population of students (not a sample), you'd divide by 5 instead: 256 ÷ 5 = 51.2, giving σ ≈ 7.16—slightly smaller because you're not adjusting for sampling error.
Common mistakes to avoid
Forgetting Bessel's correction: Many people mistakenly use n instead of n−1 when analyzing sample data. This underestimates true variability. Only use n if your data genuinely represents the complete population.
Confusing standard deviation with mean absolute deviation: Standard deviation weights larger deviations more heavily (because differences are squared), while mean absolute deviation treats all deviations equally. They answer slightly different questions about spread.
Misinterpreting units: Standard deviation is always in the same units as your original data. If measuring height in centimeters, standard deviation is also in centimeters—not squared. Variance, however, is in squared centimeters, which is why it's less intuitive for practical communication.
Ignoring outliers: Extreme values disproportionately inflate standard deviation because differences are squared. Always examine your data visually or check for data-entry errors before finalizing results.