From Variance to Standard Deviation
In the previous lesson, we learned about variance - a measure of how spread out values are. Variance is powerful, but it has a practical problem: it is measured in squared units. If you are looking at test scores measured in points, the variance comes out in "points squared." That is hard to interpret.
Standard deviation solves this by taking the square root of the variance. This brings the measurement back into the original units - points, dollars, degrees, kilograms, or whatever you started with.
Standard deviation = √variance
From our salary example in the previous lesson, the variance was 125,000 (in squared dollars).
Standard deviation = √125,000 ≈ $354
Now we can say: on average, each employee's salary differs from the mean by about $354. That is a number we can actually understand and use.
What Standard Deviation Really Means
Think of standard deviation as the average distance from the average. It is not technically an exact average of distances (the math involves squaring and square-rooting), but that intuition is very close and extremely useful.
A small standard deviation means the data points are huddled close to the mean. A large standard deviation means they are spread far apart.
Two coffee shops track how many minutes customers wait for their order over 5 visits:
Shop A: 3, 4, 4, 5, 4 → Mean = 4 minutes, Std Dev ≈ 0.6 minutes
Shop B: 1, 2, 4, 6, 7 → Mean = 4 minutes, Std Dev ≈ 2.3 minutes
Both shops have the same average wait time, but Shop A is far more consistent. If you hate unpredictable waits, Shop A is your place.
Calculating Standard Deviation Step by Step
Let us work through a full calculation so you can see exactly how it works. Do not worry about memorizing the formula - understanding the idea is what matters.
A student's quiz scores over 6 weeks: 70, 80, 75, 85, 90, 80
Step 1 - Find the mean:
(70 + 80 + 75 + 85 + 90 + 80) ÷ 6 = 480 ÷ 6 = 80
Step 2 - Find each score's distance from the mean:
- 70 − 80 = −10
- 80 − 80 = 0
- 75 − 80 = −5
- 85 − 80 = +5
- 90 − 80 = +10
- 80 − 80 = 0
Step 3 - Square each distance: 100, 0, 25, 25, 100, 0
Step 4 - Find the mean of the squared distances (variance):
(100 + 0 + 25 + 25 + 100 + 0) ÷ 6 = 250 ÷ 6 ≈ 41.7
Step 5 - Take the square root: √41.7 ≈ 6.5
The standard deviation is about 6.5 points. This means the student's scores typically land about 6 or 7 points away from their average of 80.
The 68-95-99.7 Rule
When data follows a bell-shaped pattern (which many real-world measurements do - heights, blood pressure, test scores), there is a remarkably consistent pattern in how the data distributes around the mean. This is called the 68-95-99.7 rule, sometimes called the empirical rule.
- About 68% of values fall within 1 standard deviation of the mean
- About 95% of values fall within 2 standard deviations of the mean
- About 99.7% of values fall within 3 standard deviations of the mean
Let us see what this looks like with a real example.
Suppose exam scores in a large class have a mean of 75 and a standard deviation of 10.
68% of students scored between 65 and 85 (75 ± 10)
95% of students scored between 55 and 95 (75 ± 20)
99.7% of students scored between 45 and 105 (75 ± 30)
If someone scored 50, they are more than 2 standard deviations below the mean - only about 2.5% of students did worse. That is a very low score relative to the class.
This rule is incredibly powerful because it lets you quickly judge whether a particular value is ordinary or unusual, without complicated calculations.
Interpreting Standard Deviation in Real Life
Health: Blood Pressure Readings
A doctor tells you that normal systolic blood pressure has a mean of about 120 mmHg with a standard deviation of about 15. Using the 68-95-99.7 rule, you know that most people fall between 105 and 135. A reading of 160 is almost 3 standard deviations above the mean - that is unusual and worth investigating.
Manufacturing: Quality Control
A factory produces bolts that should be 10.0 cm long. The manufacturing process has a standard deviation of 0.02 cm. This means 95% of bolts are between 9.96 cm and 10.04 cm. If a bolt measures 10.1 cm - that is 5 standard deviations away from the target - something went wrong with the machine.
Finance: Stock Market Volatility
When financial analysts talk about "volatility," they are often talking about standard deviation. A stock with a standard deviation of 2% in daily returns is relatively stable. A stock with a standard deviation of 8% is a wild ride. This helps investors match their investments to how much risk they can handle.
Two investment funds both returned an average of 7% per year over the past decade.
Fund A: Standard deviation of 3% → In most years, returns were between 4% and 10%.
Fund B: Standard deviation of 12% → In some years returns were 19%, in others they were −5%.
A retiree who needs stable income would likely prefer Fund A. A younger investor with decades ahead might accept Fund B's ups and downs for the chance of occasional high returns.
Common Misunderstandings
"A large standard deviation means the data is bad"
Not necessarily. Some things naturally vary a lot. Daily temperatures in a desert have a large standard deviation because deserts genuinely swing between extreme heat and cold. The data is not "bad" - it is accurately reflecting reality.
"Standard deviation only works for bell-shaped data"
The 68-95-99.7 rule specifically applies to bell-shaped (normal) distributions. But standard deviation itself can be calculated for any dataset. It is always a useful measure of spread, even if the exact percentages from the empirical rule do not apply perfectly.
"You need to memorize the formula"
In practice, calculators, spreadsheets, and software compute standard deviation instantly. What matters is understanding what the number means and how to use it. If someone tells you the standard deviation is 5, you should know that most values are within about 5 units of the average - that is the essential insight.
Standard Deviation at a Glance
Here is a quick way to think about different sizes of standard deviation relative to the mean:
- Small (relative to the mean): Data is tightly packed. High consistency. Think of a precise thermometer or a reliable bus schedule.
- Moderate: Normal amount of variation. Most real-world data falls here.
- Large (relative to the mean): Data is widely spread. High variability. Think of startup revenues or daily step counts that swing from 500 to 15,000.
Standard deviation tells you, in plain units, how far values typically sit from the average. A small standard deviation means consistency; a large one means wide variation. The 68-95-99.7 rule gives you a quick way to judge whether a particular value is normal or unusual: about two-thirds of values fall within one standard deviation of the mean, and nearly all fall within three. This single number is one of the most useful tools in all of statistics.