DAO1704 Chap.4 Descriptive Measures and Distribution Traps
Descriptive Measures and Distribution Traps
One average is not a distribution
The mean uses every value and responds to extremes, while the median follows order and resists them. Range uses only the minimum and maximum; interquartile range describes the middle half; standard deviation measures dispersion around the mean.
A defensible summary pairs centre with spread and follows the observed shape.
Outlier is a prompt to investigate
The boxplot fence identifies values far from the central half, but it does not decide whether they are errors. A flagged point may be a unit mistake, a real rare event or the case that matters most to a capacity decision.
Validate its source, report sensitivity and avoid deleting it merely because the mean becomes easier to explain.
The decision chooses the statistic
Median may describe a typical day, mean may support expected total demand and an upper percentile may guide surge planning.
The chapter connects each statistic to the cost of being wrong, then closes with a sensitivity statement showing whether the recommendation changes under another defensible summary or after a flagged observation is verified.
Precision should match the evidence
Extra decimal places do not solve uncertainty about membership, units or the statistic's purpose.
Report enough precision for the decision, preserve the original measurement scale and distinguish a verified correction from an analytical exclusion. When one observation controls the conclusion, show the result both ways and explain which reading governs. This sensitivity statement lets the reader see whether the recommendation is stable, while a single rounded average would hide the dependency.
What this chapter covers
- 01
Mean, median and mode as centres
- 02
Range, MAD, variance and standard deviation
- 03
Quartiles and interquartile range
- 04
Boxplot fences and source validation
- 05
Skew and robust summaries
- 06
Sensitivity statements for decisions
Worked example · free
Describing a right-skewed week
- 2Calculate the mean and median.
- 2Find Q1, Q3 and the IQR.
- 2Apply the upper boxplot fence.
- 2Interpret typical and total demand separately.
Key terms
- Mean
- The total of numerical observations divided by their count.
- Median
- The middle ordered value, or the midpoint of the two middle values.
- Standard Deviation
- The square root of variance, expressing spread around the mean in the original units.
- Interquartile Range
- The distance between the third and first quartiles.
- Right Skew
- A distribution with a longer tail toward larger values.
- Sensitivity Analysis
- A check of whether a conclusion changes under another defensible input or method.
Descriptive Measures and Distribution Traps FAQ
When should the median be preferred to the mean?
Prefer the median when the decision concerns a typical observation and strong skew or extreme values pull the mean away from most cases. Keep the mean when totals or additive planning matter, and explain the difference rather than replacing one silently.
Does a boxplot outlier have to be removed?
No. The fence is a screening device, not an error detector. Check the original record, units and event context. If the point is genuine, retain it and show whether the conclusion changes with and without it.
Why report spread beside a centre?
Groups can share the same mean or median while having very different consistency and tail risk. Spread shows how representative the centre is and helps connect the summary to service levels, capacity or uncertainty rather than presenting a single polished number.
Exam move
Take three small datasets with the same mean but different shapes. Calculate mean, median, range, IQR and standard deviation, then draw a histogram and boxplot. For each dataset, write which number answers a typical-case question, which supports a total-demand question and which warns about risk. Change or verify one extreme value and record whether the recommendation moves. This creates a reusable sensitivity habit.
Add a short decision context to each dataset and choose the cost of underestimating or overestimating. Revisit the statistic after naming that cost. You should be able to defend why the measure fits the decision rather than saying only that it is robust or familiar. Practise stating the result twice: once for an ordinary operating day and once for surge planning.
The different sentences should reveal why one centre cannot carry both decisions.
Working through Descriptive Measures and Distribution Traps in DAO1704? Sia is AskSia’s AI Statistics tutor — ask any DAO1704 Descriptive Measures and Distribution Traps question and get a clear, step-by-step explanation grounded in how DAO1704 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
