01 · A negative minimum
Find the range of −8, −3, 1, 5, 7.
Hint
Subtract −8 from 7.
Worked solution
7 − (−8) = 15.
Understand · explore · practise
Compare range, interquartile range and interpercentile ranges, including extreme values, grouped estimates and cautious contextual interpretation.
Before you startQuartiles, percentiles and grouped interpolation.
01 / Separate centre from spread
Range = maximum − minimum. IQR = Q₃ − Q₁.
Both are differences in the original variable’s units.
The range uses the two extremes. The IQR measures the distance between the lower and upper quartiles. Always state which measure you are comparing.
Values: 2, 4, 6, 8, 10, 12, 14, 20.
Range = 20 − 2 = 18.
Q₁ = (4 + 6)/2 = 5; Q₃ = (12 + 14)/2 = 13.
IQR = 13 − 5 = 8. Median = 9.
Here the largest value changes without changing the ranks used for the quartiles. That explains this example’s unchanged IQR; it is not a universal guarantee for every list.
02 / Use the observed extremes
Minimum = −4; maximum = 6
Negative values still have an ordinary ordering.
Range = 6 − (−4) = 10 °C
Subtract the minimum, including its sign.
The observed interval is −4 to 6 °C
That interval is not the numerical range.
Find the range of −8, −3, 1, 5, 7.
Subtract −8 from 7.
7 − (−8) = 15.
Values 0, 2, 4, 9 have frequencies 0, 3, 2, 1. Find the observed range.
Zero does not occur.
The minimum observation is 2, not 0; maximum 9. Range = 7.
03 / Subtract quartile values
Q₁ = (4 + 6)/2 = 5; Q₃ = (12 + 14)/2 = 13
Use this course’s stated listed-data convention.
IQR = 13 − 5 = 8
Compare values, not ranks.
Median = (8 + 10)/2 = 9
The centre is a separate summary.
Journey times have Q₁ = 18 and Q₃ = 31 minutes. Find and interpret the IQR.
Take the difference.
IQR = 13 minutes: the interval between the two quartile values is 13 minutes wide.
For 1, 3, 4, 6, 7, 8, 10, 12, 14, 19, use this course’s rule to find IQR.
Q₁ is rank 3; Q₃ is rank 8.
Q₁ = 4, Q₃ = 12; IQR = 8.
04 / Why IQR is often more resistant
Pause, replay or seek freely. The notes explain the same idea and stay in view.
In the model, raising the maximum changes the range directly, while the quartile positions continue to use unchanged values. The IQR is often more resistant to extremes than the range. However, quartiles can change if edited values enter their positions or if the list is small.
Raise the model maximum from 20 to 100. State the old and new range and IQR.
The minimum stays 2; quartile values stay 5 and 13.
Range rises from 18 to 98. IQR remains 8 in this particular list.
Under this lesson’s quartile rule, compare the IQRs of 0, 1, 2, 3 and 0, 1, 2, 100.
For four values, average ranks 1–2 for Q₁ and 3–4 for Q₃.
First: Q₁ = 0.5, Q₃ = 2.5, IQR = 2. Second: Q₁ = 0.5, Q₃ = 51, IQR = 50.5. Thus “IQR never changes when an extreme changes” is false.
05 / Choose a wider central interval
The interdecile range is P₉₀ − P₁₀. More generally, an interpercentile range is the upper percentile value minus the lower percentile value. The 10th-to-90th interval summarises the central 80% in a continuous model, but tied finite data need not place exactly 80% strictly inside the endpoints.
A summary gives P₁₀ = 8 and P₉₀ = 35 seconds. Calculate the interdecile range.
Subtract the tenth percentile value.
35 − 8 = 27 seconds.
P₅ = 2.5 and P₉₅ = 17.2 kg. Give the interpercentile range and the central percentage it targets.
Use the percentile labels for percentage; values for width.
Width = 17.2 − 2.5 = 14.7 kg. The interval targets the central 95 − 5 = 90% in a continuous distribution.
06 / Retain the uncertainty of grouping
For grouped data, interpolate each quartile using its own class and then subtract. The exact sample range is usually unknown: class boundaries are not necessarily observed extremes. A boundary span can give a useful upper bound or a rough range estimate, but must be labelled.
Interpolated quartiles are 11.25 and 27.5 minutes. Find the estimated IQR.
Subtract the estimates.
27.5 − 11.25 = 16.25 minutes. It remains an estimate based on the within-class interpolation assumption.
The lowest occupied interval is [10,20) and the highest is [40,50). Is the observed range exactly 40?
The minimum is at least 10 and the maximum is below 50.
No. The boundary span is 50 − 10 = 40, but the actual range is less than 40. Since a lowest-class observation is below 20 and a highest-class observation is at least 40, the actual range is greater than 20. Thus 20 < range < 40; the exact value is unknown.
07 / Compare like with like
Service A has median journey time 22 minutes and IQR 6; B has median 19 and IQR 11. Compare the summaries.
Discuss centre and spread separately.
B has the lower median journey time, while A has the smaller spread between quartiles. These summaries do not show every journey or establish that one service causes a particular outcome.
Dataset A has IQR 0.2 hours and B has IQR 10 minutes. Which IQR is smaller?
Convert to a common unit.
A: 0.2×60 = 12 minutes. B: 10 minutes, so B has the smaller IQR. Comparing 0.2 with 10 without units would mislead.
08 / State the measure and its limits
Report the numerical difference with units, say whether it is exact or estimated, and explain what it describes. Check quartile conventions and measurement units before comparing groups.
Section 1 of 8 · Separate centre from spread