The range of a dataset is its largest value minus its smallest value — one subtraction, and you know how far the data stretches end to end. If you’re trying to figure out how to find the range for a set of numbers right in front of you, that subtraction is the entire calculation.
What’s worth understanding beyond that one line is when the range tells you something real and when it quietly misleads you. A single unusual reading can make a tightly clustered dataset look wildly spread out, and that failure mode catches people who lean on the range without checking it. This guide walks through the calculation on a real dataset, covers the cases that trip people up — negative numbers, decimals, grouped data — and shows exactly where the range stops being trustworthy and what to check instead.
What the Range Measures
The range is the simplest of the variability statistics because it only looks at two numbers in the whole dataset: the maximum and the minimum. Every other value — everything in between — is ignored.
That’s also its biggest limitation. The mean, variance, and standard deviation are built from every data point; the range is built from two. It sits alongside the mean, median, mode, variance, and standard deviation in the standard descriptive-statistics toolkit, but it earns its place there by being the fastest of them to compute by hand — whether the data is raw, negative, decimal, or already grouped into a frequency table, the underlying operation never changes.
The formula:
Range = Maximum value − Minimum value
Range = x_max − x_min
Where x_max is the largest value in the dataset and x_min is the smallest. You don’t need the mean, the median, or a running total — just the two extremes.
Finding the Range: A Three-Step Process
Calculating the range never takes more than three moves.
Step 1 — Gather every value
List out the dataset. If it’s already sorted, you’re a step ahead. If not, don’t bother sorting the whole thing just to find the range — scanning once for the two extremes is faster than a full sort.
Step 2 — Identify the maximum and minimum
Scan the list and record the largest value (x_max) and the smallest value (x_min). For a short list you can usually spot both by eye. For a long one, sorting first makes them impossible to miss.
Step 3 — Subtract the minimum from the maximum
Range = x_max − x_min
The result should always come out zero or positive. If you land on a negative number, the subtraction happened backwards.
Worked Example: Patient Wait Times at an Urgent Care Clinic
A clinic manager wants to know how consistent patient wait times were on a Tuesday shift, before deciding whether the clinic needs a second triage nurse. Over one 12-patient shift, wait times from check-in to being seen (in minutes) were:
8, 15, 22, 9, 12, 6, 18, 17, 14, 13, 11, 19
Step 1 — List the values: all 12 are recorded above.
Step 2 — Find the max and min. Sorted ascending, the extremes are easy to spot:
6, 8, 9, 11, 12, 13, 14, 15, 17, 18, 19, 22
Minimum: 6 minutes. Maximum: 22 minutes.
Step 3 — Subtract:
Range = 22 − 6 = 16 minutes
That’s how to find the range of this shift’s data: it’s 16 minutes. On its own, that number tells the manager patients weren’t waiting wildly different lengths of time — but it says nothing about whether most people waited close to 6 minutes or close to 22. Keep this dataset in mind; it’s coming back later, because the gap between “spread” and “typical” is exactly where the range can mislead you.
Try the Range Calculator
Enter your own set of numbers and the calculator finds the range, minimum, and maximum instantly.
For a full-page version, open the dedicated range calculator, or browse the full calculators hub.
Special Cases: Negative Numbers, Decimals, and Grouped Data
The three-step process doesn’t change for messier data, but three situations trip people up often enough to walk through on their own.
Negative Numbers
Dataset: overnight low temperatures (°C) recorded at a weather station over one week: −12, 3, −7, 0, −18, 5, −4.
Sorted: −18, −12, −7, −4, 0, 3, 5.
Maximum: 5. Minimum: −18.
Range = 5 − (−18) = 5 + 18 = 23
The range is 23°C. Subtracting a negative number is the same as adding its absolute value — the mistake to watch for is treating −18 itself as “the range,” or forgetting to flip the sign. Write the subtraction out in full rather than doing it in your head.
Decimal Values
Nothing changes except lining up the decimal point correctly.
Dataset: body-mass index (BMI) readings from six patients at a routine screening: 22.4, 27.1, 19.8, 31.5, 24.6, 18.3.
Maximum: 31.5. Minimum: 18.3.
Range = 31.5 − 18.3 = 13.2
The BMI range across this group is 13.2.
Grouped Data (Frequency Tables)
When data arrives already binned into class intervals, you don’t have the individual values anymore — so the range becomes an estimate built from the boundaries of the two end classes, not an exact figure.
Example: a regional courier logs weekly delivery times (minutes) in bins for an operations report:
| Class interval (minutes) | Deliveries |
|---|---|
| 20–30 | 6 |
| 30–40 | 14 |
| 40–50 | 9 |
| 50–60 | 3 |
The lower boundary of the lowest class is 20; the upper boundary of the highest class is 60.
Range ≈ 60 − 20 = 40 minutes
This is an upper-bound estimate, not a measured value — the actual fastest and slowest deliveries inside those two end classes are unknown, so the true range could be smaller. Report it as “at most 40 minutes,” not as an exact figure.
Range vs. IQR and Standard Deviation: When Each One Misleads You
Go back to the urgent-care example: 12 patients, wait times from 6 to 22 minutes, a range of 16. Now suppose a 13th patient on that same shift needed a complex work-up and waited 95 minutes for a specialist consult. None of the other 12 patients’ wait times changed — but recompute the range and it jumps to 89 minutes (95 − 6), more than five times larger, because of a single additional value.
That’s the range’s core weakness: it’s built from exactly two numbers, so one unusual reading anywhere in the dataset can dominate the result. The open-access textbook OpenStax, Introductory Statistics — 2.7 Measures of the Spread of the Data covers the range alongside the other spread measures, which is useful for seeing how they’re meant to complement each other rather than substitute for one another.
Range vs. Interquartile Range (IQR)
The interquartile range trims the extremes away before measuring anything — it’s the distance between the 75th percentile (Q3) and the 25th percentile (Q1), so the top and bottom quarters of the data can’t influence it.
IQR = Q3 − Q1
Run both measures on the clinic data: before the 95-minute outlier, IQR = 7.5 minutes; after adding it, IQR = 8.5 minutes — it barely moved, because the outlier landed outside the middle 50% the IQR actually measures. The range, over that same comparison, moved from 16 to 89. Whenever a dataset might contain a data-entry error or a genuine one-off event, the IQR is the more trustworthy spread statistic to report.
Range vs. Standard Deviation
Standard deviation measures the average distance of every point from the mean, not just the two extremes, so it reflects the whole shape of the data at the cost of a heavier calculation. A rough rule of thumb: in a large, roughly normal dataset, the range tends to run about four to six standard deviations wide (4σ to 6σ). If your computed range sits far outside that band, treat it as a signal to check for an outlier or a skewed distribution before trusting either number.
A Quick Comparison
| Measure | Uses all values? | Resistant to outliers? | Units | Effort |
|---|---|---|---|---|
| Range | No — only 2 | No | Same as data | Lowest |
| IQR | No — middle 50% | Yes | Same as data | Low |
| Variance | Yes | No | Squared | Moderate |
| Standard deviation | Yes | No | Same as data | Moderate |
Reach for the range when you want a fast, easy-to-explain first pass, when you’re scanning for data-entry errors (an implausibly large range is often a typo), or when you’re picking histogram bin widths. Reach for the IQR or standard deviation the moment you suspect outliers, or when “spread” needs to describe what’s typical rather than what’s extreme.
Common Mistakes When Calculating the Range
Subtracting in the wrong order
The range is always maximum minus minimum. Reverse the order and you get a negative number that means nothing as a measure of spread — if your result comes out negative, you subtracted backwards.
Confusing the range statistic with the range interval
“The data ranges from 6 to 22” describes an interval; the range itself is a single number — in that example, 16. Keep the two uses of the word separate when you’re reporting results, especially in writing.
Forgetting the sign rule with negative numbers
Subtracting a negative adds its absolute value: 5 − (−18) equals 23, not 5 − 18 = −13. Write the subtraction out in full rather than doing the arithmetic in your head.
Treating “the range” and “the spread” as interchangeable
After a shift, an exam, or a batch of measurements, it’s common to hear “the range was high” when the speaker actually means variance or standard deviation was high. They aren’t the same number — and, as the clinic example above shows, the range can spike from a single outlier while every other measure of spread stays exactly where it was.
Frequently Asked Questions
How do I calculate range when the dataset has negative numbers?
Exactly as for positive numbers: find the maximum and minimum as usual, then subtract. Because you’ll typically be subtracting a negative minimum, remember that subtracting a negative number adds its absolute value — for example, if x_max = 5 and x_min = −18, Range = 5 − (−18) = 23.
How do you calculate for range in a grouped frequency table?
Use the upper boundary of the highest class interval minus the lower boundary of the lowest class interval. This produces an approximation, since the individual values inside each class aren’t available — the true range could be somewhat smaller than this estimate.
When should I use the range instead of the standard deviation?
Use the range for a quick, easy-to-explain snapshot of spread, especially in a small dataset without outliers. Use the standard deviation when you need a fuller picture — one that reflects how every data point varies from the mean, not just the two extremes. When outliers are a real possibility, the IQR is usually the better choice over both.
What counts as a “good” range for a dataset?
There’s no universal “good” range — it depends entirely on context. A range of 5 points is tight for exam scores out of 100 and enormous for precision manufacturing tolerances measured in micrometres. Judge the range relative to the scale of the measurement and the purpose of the analysis, not against some fixed number.
Summary
The range is one subtraction: find the maximum, find the minimum, subtract.
Range = Maximum − Minimum
It’s the fastest variability statistic to compute and the easiest to explain to a general audience, and it works identically for positive numbers, negative numbers, decimals, and grouped-data estimates. Its one real weakness is sensitivity to outliers — as the urgent-care example showed, a single unusual value can multiply the range several times over while the rest of the data doesn’t move at all. When that risk is live, pair the range with the IQR or the standard deviation rather than reporting it alone.
Use the range for a quick first-pass check, for setting histogram bin widths, or for reporting spread to someone who doesn’t need the full statistical picture. For a closer look at how spread is calculated and notated more rigorously, see the standard deviation symbol guide, or browse the full calculators hub.
For a standards-body reference on measures of scale, see the NIST/SEMATECH e-Handbook of Statistical Methods, Section 1.3.5 — Measures of Scale.