Standard Deviation Explained Simply (With Steps)

Standard deviation is the number statistics classes rush past and everyone else quietly forgets. That's a shame, because it answers a genuinely useful question: on average, how far are the data points from the average? Low standard deviation means the data huddles around the mean; high standard deviation means it spreads out.

The 5-step calculation

Take a small dataset — five quiz scores: 72, 78, 80, 85, 90.

Step 1 — Find the mean. (72 + 78 + 80 + 85 + 90) / 5 = 405 / 5 = 81

Step 2 — Subtract the mean from each value (deviation).

Step 3 — Square each deviation. (Squaring removes negative signs and penalizes large gaps more than small ones.)

Step 4 — Average the squared deviations (variance).

Step 5 — Take the square root (standard deviation).

Variance = 188 / 5 = 37.6 Standard deviation = √37.6 ≈ 6.13

So the scores land, on average, about 6 points from the mean of 81. Confirm the number with the Standard Deviation Calculator, and use the Mean, Median, Mode Calculator if you want the center measures for the same list.

Population vs sample (n vs n−1)

The example above used the population formula — dividing by n = 5. If those five scores were a *sample* drawn from a larger class, you'd divide by n − 1 = 4 instead:

Sample variance = 188 / 4 = 47 Sample SD = √47 ≈ 6.86

Why n − 1? Sample means underestimate the true population spread slightly, and dividing by n − 1 (Bessel's correction) compensates. Rule of thumb: use n − 1 when your data is a sample; n when you have the entire population. Statistical software defaults to n − 1.

The 68-95-99.7 rule

For data that's roughly bell-shaped (normally distributed):

Applied to our quiz: if scores are normally distributed with mean 81 and SD ~6, you'd expect about 68% of students between 75 and 87, and virtually all between 63 and 99. That's why SD is central to z-scores — a z-score is just "how many standard deviations from the mean is this value?"

When SD is (and isn't) the right measure

SD is a good spread measure when the data is roughly symmetric and free of extreme outliers. When distributions are skewed — house prices, income, wait times — SD gets pulled by the long tail and the interquartile range (IQR) or median absolute deviation is often more informative.

Two datasets can share a mean of 81 and be nothing alike:

• A: 80, 81, 81, 81, 82 → SD ≈ 0.71
• B: 55, 70, 81, 92, 107 → SD ≈ 20.6

Both average 81. The first is a class where everyone learned the material; the second is a class you need to think about differently.

FAQ

What does standard deviation actually measure? The typical distance between individual data points and the mean. Small SD = tightly clustered data; large SD = widely spread data.

When do I use n vs n−1 for standard deviation? Use n when you have every data point in the population. Use n − 1 when your data is a sample drawn from a larger population — the correction avoids underestimating true variability.

What is a "good" standard deviation? There isn't one — SD is only meaningful relative to the mean and the context. A ±5 point SD on quiz scores means one thing; a ±5% SD on a stock return means another. Compare like to like.

---

*Skip the arithmetic and run your list through the free Standard Deviation Calculator — supports both population and sample formulas.*