Two data sets can have the same mean but very different consistency. Standard deviation measures how far observations typically spread from the mean, using squared deviations so positive and negative departures do not cancel.
From deviations to standard deviation
For a population of N values with mean variance is the average squared deviation, and standard deviation is its square root:
The square root restores the original units. If observations are measured in centimeters, variance has squared-centimeter units while standard deviation is in centimeters.
For a sample used to estimate population spread, the usual sample standard deviation s divides the squared-deviation sum by instead of . The sample mean has already been estimated from the data, which uses one degree of freedom. Follow whether the question asks for population or sample standard deviation.
Example: see spread around the same center
The sets 4, 5, 6 and 1, 5, 9 both have mean 5. The first has deviations , 0, 1; the second has deviations , 0, 4. The second set is more spread out even before any formula is evaluated.
Their population standard deviations are and The second is four times the first because every deviation from the common mean is four times as large.
A standard deviation of zero means all observations are equal. It does not merely mean the mean is zero. For example, and 2 have mean zero but positive spread.
Example: calculate a population standard deviation
For 2, 4, 4, 4, 5, 5, 7, 9, the mean is 5. The squared deviations are 9, 1, 1, 1, 0, 0, 4, and 16, with sum 32. Divide by 8 to get variance 4, then take the square root to get standard deviation 2.
The sample standard deviation of the same eight observations would be not 2. Calculators often show both values; the notation or context determines which to use.
You do not need to calculate standard deviation to compare obvious equal-center spreads. Numerical work should support the conceptual interpretation, not replace it.
Dots at 2, three at 4, two at 5, one at 7, and one at 9. The mean is 5.
The data 2, 4, 4, 4, 5, 5, 7, 9 are centered at 5, but their distances from 5 vary.
Describe this figure
Dots at 2, three at 4, two at 5, one at 7, and one at 9. The mean is 5.
Values: 2, 4, 4, 4, 5, 5, 7, 9.
Adding shifts center; multiplying changes spread
Adding c to every observation adds c to the mean and median but leaves standard deviation, range, and IQR unchanged. Every point and the mean move together, so deviations stay the same.
Multiplying every observation by multiplies the mean by and standard deviation by Variance is multiplied by The absolute value is needed because spread cannot be negative.
For a temperature conversion a Celsius standard deviation of 5 becomes 9 Fahrenheit degrees. The added 32 changes the mean but contributes nothing to spread.
Extreme values and resistant summaries
An extreme observation can strongly affect the mean and standard deviation because its squared deviation is large. Median and IQR are more resistant to the magnitude of extreme observations, so they are often preferred for skewed distributions or data with outliers.
The 1.5-IQR rule flags values below or above It is a screening rule, not a reason to delete data automatically. Check the context and whether an observation is an error or a real unusual case.
Try it yourself
A data set has mean 12 and standard deviation 3. Each value is transformed by Find the new mean and standard deviation.
Answer: The new mean is The new standard deviation is Subtracting 7 shifts all values equally and does not reduce spread.
What to remember
Standard deviation describes spread around the mean in the original units. Distinguish sample from population formulas, and separate additive shifts from multiplicative changes.