Select Page

Range is the largest value in a data set minus the smallest value. It uses the same units as the data, so a range of 0.08 mm means the data spans 0.08 mm. Six Sigma teams use range in two main ways. First, they plot the range of each small subgroup on an R chart to watch process spread over time.

Second, they divide the average range by a constant called d2 to estimate the process standard deviation. The range works well for subgroups up to about 10 observations. It ignores every value between the extremes, so one outlier can distort it.

Quick-Reference Table: Range Terms and Formulas

TermFormulaMeaning
RangeR = maximum – minimumSpread of one data set
Subgroup rangeRi = maximum – minimum within subgroup iSpread inside one small sample
Average range (R-bar)R-bar = sum of Ri / number of subgroupsTypical subgroup spread
R chart upper limitUCL = D4 x R-barUpper boundary for normal spread
R chart lower limitLCL = D3 x R-barLower boundary for normal spread
Sigma estimatesigma-hat = R-bar / d2Process standard deviation from ranges
Moving rangeMRi = absolute value of (xi – x(i-1))Spread between consecutive single values
Interquartile rangeIQR = Q3 – Q1Spread of the middle half of the data

Key Takeaways

  • Range uses only two values. It reports the maximum and the minimum, and nothing in between.
  • Range grows with sample size. A bigger sample has more chances to catch an extreme value, so do not compare ranges across different sample sizes.
  • R charts track subgroup range. They show whether process spread stays stable over time.
  • R-bar divided by d2 estimates sigma. This is how teams get a standard deviation from small subgroups.
  • Range suits small subgroups. Switch to standard deviation or an S chart for larger subgroups.
  • Range is not tolerance. Range describes data you collected. Tolerance describes what the customer allows.

What is Range in Statistics?

Range measures how far a data set stretches from its lowest to its highest value. The formula is simple.

R = maximum – minimum

Here is a small process example. Five orders took 14, 18, 15, 22, and 16 minutes to process. The maximum is 22 and the minimum is 14. The range is 22 minus 14, or 8 minutes. The answer carries the same unit as the data.

Two Dot Plots Showing Range 10
Two Dot Plots Showing Range 10

For a beginner walkthrough with another example, read our blog article on range as a measure of dispersion.

Kevin Clay

Public, Onsite, Virtual, and Online Six Sigma Certification Training!

  • We are accredited by the IASSC.
  • Live Public Training at 52 Sites.
  • Live Virtual Training.
  • Onsite Training (at your organization).
  • Interactive Online (self-paced) training,

What Range Tells You and What It Does Not?

Range is a fast first look at spread. It also leaves a lot out.

Range Tells YouRange Does Not Tell You
How far apart the extremes areHow the middle values are arranged
A quick sense of variationWhether one outlier caused the spread
Spread in the data’s own unitsHow many values sit near the extremes

Why Two Data Sets Can Share a Range and Still Differ

Two data sets can have the same range and different spread. Compare these two sets of five values.

Data SetValuesRangeStandard Deviation
A10, 10, 10, 10, 20104.47
B10, 12, 14, 16, 20103.85

Both sets have a range of 10. Set A has four identical values and one far-off point. Set B spreads evenly. The standard deviations differ because that statistic uses every value. Range cannot see the difference.

Also Read: I-MR Chart: Mastering Individual Moving Range Control Charts

How One Outlier Distorts Range?

A single extreme value can swing the range. Take this data set: 10, 11, 11, 12, 12, 13, 13, 14, and 40.

Data SetRangeInterquartile Range
Without the 4042
With the 40302.5

The interquartile range uses the median-of-halves method. Software packages can use slightly different quartile methods.

One outlier moved the range from 4 to 30. The interquartile range moved only from 2 to 2.5. That is why analysts often check the IQR, or investigate the outlier, before they trust a range.

Why Range Grows With Sample Size?

A larger sample tends to produce a larger range. More observations give you more chances to catch an extreme value. This matters when you compare data sets of different sizes.

The effect has a name in control chart theory. If r is the range of N observations from a normal distribution with standard deviation sigma, the expected range equals d2(N) times sigma. The constant d2 rises as N rises.

Sample Size (n)Expected Range (in units of sigma)
21.128
52.326
103.078

For n = 2 the average range is 1.128 sigma, and for n = 5 it is 2.326 sigma. Practical rule: compare ranges only when the sample sizes match.

Where Range Appears in a Six Sigma Project

Range shows up in several places, even when the word is not in the title.

WherePhaseHow Range Helps
Quick data checkMeasureScreens spread before heavy analysis
Process capabilityMeasureR-bar divided by d2 estimates within-subgroup sigma
Gage R&RMeasureA classic method uses average ranges of repeated measurements
Multi-vari studyAnalyzeRange within one part shows positional variation
R chartControlTracks subgroup spread over time
Moving range chartControlTracks spread for single readings

Our glossary entries on Multi-Vari analysis and precision cover two of these uses in detail.

Also Read: Green Belt Syllabus: Every Topic in SSDSI’s Online Course

How an R Chart Works

An R chart plots the range of each subgroup in time order. A subgroup is a small sample taken under the same conditions, such as five consecutive parts. Follow these steps.

  1. Collect subgroups of equal size. Use the same number of readings each time.
  2. Calculate each subgroup’s range. Subtract the minimum from the maximum.
  3. Calculate R-bar. Average all the subgroup ranges.
  4. Look up D3 and D4 for your subgroup size.
  5. Calculate the limits. The upper limit equals D4 times R-bar. The lower limit equals D3 times R-bar.
  6. Plot and read the chart. Look for points outside the limits and unusual patterns.

Check the R chart before the X-bar chart. If the R chart shows that sample variability is out of control, the whole process is judged out of control regardless of what the X-bar chart shows. Our x-bar glossary entry covers the paired chart.

Control Chart Constants for Range

These constants come from standard control chart tables.

nD3D4d2
203.2671.128
302.5741.693
402.2822.059
502.1142.326
602.0042.534
70.0761.9242.704
80.1361.8642.847
90.1841.8162.970
100.2231.7773.078

Tables round slightly differently. Some list D4 as 2.115 for n = 5 and 2.575 for n = 3. Check the table your software or standard uses. One published table lists the same values for n = 2 through 10.

Why D3 is zero for small subgroups. The lower factor is set to zero when the theoretical lower limit would fall below zero. A range cannot be negative, so the table uses zero.

A Worked Example: R Chart and Sigma Estimate

This example reuses the subgroup data from our x-bar glossary entry. A line pulls five parts each hour and measures a dimension in millimeters.

SubgroupValuesRange
110.01, 9.99, 10.02, 9.98, 10.000.04
210.03, 10.01, 9.97, 10.00, 9.990.06
39.95, 10.02, 10.01, 9.99, 10.030.08
410.02, 10.00, 9.98, 10.01, 9.990.04

Step 1: Find R-bar. Add the ranges: 0.04 plus 0.06 plus 0.08 plus 0.04 equals 0.22. Divide by 4 to get R-bar of 0.055 mm.

Step 2: Find the R chart limits. For n = 5, D4 is 2.114 and D3 is 0.
UCL = 2.114 x 0.055 = 0.116 mm
LCL = 0 x 0.055 = 0 mm

A subgroup range above 0.116 mm signals unusual spread worth investigating.

Step 3: Estimate sigma. For n = 5, d2 is 2.326.
sigma-hat = 0.055 / 2.326 = 0.0236 mm

Step 4: Compare spread with tolerance (illustration). Suppose the drawing allows 10.00 plus or minus 0.10 mm, a tolerance width of 0.20 mm. The process spread is 6 times 0.0236, or about 0.142 mm. A quick Cp is 0.20 divided by 0.142, or about 1.41. Cp ignores centering, so check Cpk too.

A caution. Four subgroups keep the arithmetic short. Many references recommend collecting 20 to 25 or more subgroups before you trust control limits.

The Moving Range for Single Readings

Some processes give one reading per period. You cannot form subgroups, so you use the moving range. Each moving range is the absolute difference between two consecutive points.

For an individuals chart, the n = 2 constants apply: D4 is 3.267, D3 is 0, and d2 is 1.128. Here is a short example with daily readings of 50, 53, 49, 52, and 55.

PairMoving Range
50 to 533
53 to 494
49 to 523
52 to 553

MR-bar is 13 divided by 4, or 3.25. The sigma estimate is 3.25 divided by 1.128, or about 2.88. The moving range upper limit is 3.267 times 3.25, or about 10.62.

Range, Standard Deviation, and IQR: Which One to Use

MeasureUsesBest WhenWeak Spot
RangeTwo valuesSubgroups of about 10 or fewer, quick checks, R chartsOutliers and sample size
Standard deviationEvery valueLarger samples, most statistical tests, S chartsHarder to calculate by hand
Interquartile rangeMiddle half of the dataSkewed data or outliersIgnores the tails

An X-bar and R chart suits subgroup sizes of about 10 or fewer. For larger subgroups, use the standard deviation and an S chart. See our glossary entry on standard deviation.

Range, Process Spread, and Tolerance Are Different Things

These three terms get mixed up often.

TermWhat It DescribesSource
RangeSpread in the sample you collectedYour data
Process spreadAbout 6 sigma, the natural width of the processYour process
ToleranceUpper spec limit minus lower spec limitYour customer or design

A process can show a small range and still miss its tolerance if it runs off center. Compare all three before you judge capability.

Common Mistakes With Range

  • Comparing ranges from different sample sizes. The larger sample will usually show the larger range.
  • Using range for big subgroups. Past about 10 readings, standard deviation works better.
  • Ignoring outliers. Investigate extreme values before you report a range.
  • Reading the X-bar chart first. Check the R chart first, since unstable spread invalidates the average chart.
  • Treating one range as the process spread. One subgroup range is a single sample, not the whole process.
  • Confusing range with tolerance. One describes your data, and the other describes the requirement.
  • Using too few subgroups. Early limits can shift as you add data.

Frequently Asked Questions (FAQs) on Range in Statistics

What is the range in statistics?
Range is the difference between the largest and smallest values in a data set. It measures spread in the same units as the data.

How do you calculate the range?
Subtract the minimum value from the maximum value. For the values 14, 18, 15, 22, and 16, the range is 22 minus 14, which equals 8.

What is the range used for in Six Sigma?
Teams use range to build R charts, to estimate sigma from small subgroups, to run the classic average and range Gage R&R method, and to screen data for spread.

What is an R chart?
An R chart plots the range of each subgroup over time. It shows whether the spread of the process stays stable.

What is R-bar?
R-bar is the average of all the subgroup ranges. It sets the center line of the R chart and feeds the sigma estimate.

How do you estimate sigma from the range?
Divide R-bar by the constant d2 for your subgroup size. For subgroups of five, d2 is 2.326.

Why does the range increase with sample size?
A larger sample has more chances to include extreme values. The expected range rises with sample size, which is why d2 grows as n grows.

When should I use range instead of standard deviation?
Use range for small subgroups of about 10 or fewer and for quick checks. Use standard deviation for larger samples and for most statistical tests.

What is the moving range?
The moving range is the absolute difference between two consecutive single readings. It tracks spread when you cannot form subgroups.

Is range the same as tolerance?
No. Range describes the data you collected. Tolerance is the upper specification limit minus the lower specification limit, set by the customer or design.

What is the interquartile range?
The interquartile range is the third quartile minus the first quartile. It describes the middle half of the data and resists outliers.

Final Words

Range is the simplest measure of spread: the largest value minus the smallest. Its simplicity is both its strength and its limit. Six Sigma teams use it where it works best, on small subgroups, through R charts and sigma estimates built from R-bar and d2.

They avoid it where it fails, with large samples, outliers, or mismatched sample sizes. Keep range, process spread, and tolerance separate in your mind, and check the R chart before you trust any average chart.

Build Control Chart Skills With SSDSI

Reading an R chart correctly, and knowing when to switch to standard deviation, comes from practice on real data. SSDSI is IASSC accredited and delivers Green Belt and Black Belt training through onsite, live virtual, public, and online formats. The online Green Belt syllabus includes modules on basic statistics, capability analysis, and controlling the process.

SSDSI has certified 5,322+ professionals across 600+ organizations in 52 cities, backed by a 5-star Google review rating. If you want hands-on practice with these tools, a conversation about the right format is a solid next step.

About Six Sigma Development Solutions, Inc.

Six Sigma Development Solutions, Inc. offers onsite, public, and virtual Lean Six Sigma certification training. We are an Accredited Training Organization by the IASSC (International Association of Six Sigma Certification). We offer Lean Six Sigma Green Belt, Black Belt, and Yellow Belt, as well as LEAN certifications.

Book a Call and Let us know how we can help meet your training needs.