Why an average is only a starting point
Suppose two groups each have an average score of 10. In one group, nearly every score is close to 10; in the other, scores vary widely. The averages match, but they describe different experiences. Variability is the name for that spread. It helps answer a practical question: how much can individual values differ from the typical value?
Consider two small sets of scores: 8, 9, 10, 11, 12 and 6, 8, 10, 12, 14. Both have a mean of 10. The second set, however, has values farther from the mean. Looking at spread alongside the average makes that difference visible, whether you are comparing delivery times, survey responses or results from a repeated process.
The range: a quick view from smallest to largest
The range describes the distance between the lowest and highest values. Subtract the minimum from the maximum: for the first set of scores, the range is 12 minus 8, or 4. For the second, it is 14 minus 6, or 8. You can also report the endpoints themselves, such as “scores ran from 6 to 14,” which often gives readers more context than the difference alone.
A range is easy to calculate and useful for a first look at the full span of observations. Its main limitation is that it depends entirely on two values. If most delivery times are similar but one delivery takes much longer, that single result can greatly expand the range. The range does not show whether values are evenly scattered across the span or clustered in one place. It can also grow as more observations are collected, simply because a larger group offers more chances to encounter an extreme.
Standard deviation: spread around the mean
Standard deviation describes how far values typically lie from the mean, using all the values rather than just the endpoints. To calculate it, first find each value’s difference from the mean. Square those differences, average the squares using the appropriate formula, then take the square root. Squaring keeps positive and negative differences from canceling out; taking the square root returns the result to the original units.
For the two score sets above, the second has the larger standard deviation because its values sit farther from 10. A standard deviation of zero would mean every value is identical. A larger standard deviation generally indicates greater spread around the mean, but “large” needs context. A spread of two minutes may matter greatly for one process and hardly at all for another.
You may encounter slightly different standard deviation figures for the same small set of values. A calculation describing an entire group and one estimating variability from a sample use different divisors. When a report gives a standard deviation, check what data it covers and whether you are comparing figures calculated on the same basis.
Read both measures with the shape of the data
Neither measure tells the whole story. Standard deviation is closely tied to the mean, so a few unusually high or low values can affect both. Two datasets can even have similar standard deviations while looking quite different: one might cluster near the center, while another has separate groups of values. A simple plot or a list of values can reveal patterns that a single summary number hides.
Before comparing variability, check that the figures use the same units and come from comparable groups. Ask how many observations were included, whether any values are missing, and whether an extreme value represents a genuine outcome or a recording error. Do not remove an inconvenient value simply because it makes the spread look larger.
A useful reading habit is to pair three questions: What is typical? How wide is the full span? How far do values generally sit from the average? The mean, range and standard deviation offer complementary answers. Together with a look at the underlying distribution, they help you understand not just where the data centers, but how much individual outcomes may differ.
