Standard deviation generates more confusion than any other statistic. Here are the questions students and professionals ask most — with clear answers.
What does standard deviation actually mean?
Standard deviation measures the typical distance between each data point and the mean.
- σ = 0 → all values are identical
- σ = small → values cluster tightly around the mean
- σ = large → values are widely spread
Think of it as: "On average, how far off from the mean is a typical data point?"
What's the difference between population and sample SD?
| Feature | Population (σ) | Sample (s) | |---------|---------------|------------| | Data | Every member | A subset | | Divide by | n | n − 1 | | Symbol | σ (sigma) | s | | When to use | You have ALL data | You have a sample | | Result | Exact | Estimated |
Example: All 50 students in a class → use σ (divide by 50). 50 students out of 5000 → use s (divide by 49).
Why divide by n−1 instead of n?
Bessel's correction. A sample's values cluster around the sample mean, not the population mean. This makes the sample variance systematically smaller than the population variance. Dividing by n−1 compensates.
The smaller the sample, the bigger the correction:
- n = 5: dividing by 4 vs 5 → 25% difference
- n = 100: dividing by 99 vs 100 → 1% difference
- n = 1000: negligible difference
What is variance and how is it related?
Variance = σ² (standard deviation squared)
| Statistic | Units | Use | |-----------|-------|-----| | Variance | Squared units | Mathematical calculations | | Standard deviation | Original units | Interpretation and reporting |
Example: If data is in meters:
- Variance is in m² (hard to interpret)
- Std dev is in m (directly meaningful)
Always report standard deviation. Use variance for calculations.
What is a z-score?
z = (x − μ) / σ
A z-score converts any value to the number of standard deviations from the mean.
| Raw Score | μ | σ | Z-Score | Percentile | |-----------|---|---|---------|-----------| | 85 | 70 | 10 | +1.5 | 93.3% | | 60 | 70 | 10 | −1.0 | 15.9% | | 70 | 70 | 10 | 0 | 50% | | 100 | 70 | 10 | +3.0 | 99.9% |
Z-scores let you compare values from different distributions.
What is the empirical rule?
For normal distributions only:
| Within | % of Data | |--------|-----------| | μ ± 1σ | 68.27% | | μ ± 2σ | 95.45% | | μ ± 3σ | 99.73% |
Quick math: If σ = 10 and μ = 100, then about 68% of data is between 90 and 110, and about 95% is between 80 and 120.
Can standard deviation be 0?
Yes — when all values are identical. If every student scores exactly 75, then μ = 75 and σ = 0.
σ = 0 means zero variability. Every data point equals the mean.
Can standard deviation be negative?
No. Standard deviation is the square root of variance, and variance is a sum of squares (always ≥ 0). The square root of a non-negative number is always non-negative.
The Trench Truth: When someone says "the data has a standard deviation of 50," ask: "Of what?" σ = 50 is tiny for salaries in thousands but enormous for body temperature. Context matters. Always compare σ to the mean. A coefficient of variation (σ/μ) of 5% is low regardless of scale.
Related: Statistics Calculator · Derivative Calculator · Square Root Calculator
Discussion
Loading comments...