BUSS1020 Chap.2 Descriptive Measures and Visualisation
Descriptive Measures and Visualisation
A useful descriptive analysis combines centre, spread and shape. This chapter compares the mean, median and mode, then links standard deviation with the mean and the interquartile range with the median. It explains z-scores, quartiles, boxplot fences, skewness and the difference between flagging an outlier and deleting it.
Chart selection follows the variable and decision: bar charts for categories, histograms for quantitative shape, time plots for sequence and scatterplots for pairs of measurements. Equal averages can hide very different operational risk, so the worked comparison calculates variation rather than stopping at location.
Direction checks protect z-score interpretation, and a five-number summary gives a compact route from ordered data to boxplot diagnostics. The chapter also treats visual design as part of statistical honesty: truncated scales, decorative perspective and unsuitable bins can change what a reader perceives even when the underlying values are untouched.
The practice set forces every calculation to end in the original unit and asks what a flagged case means operationally. This keeps an extreme value visible as evidence to investigate instead of an inconvenience to erase. A reliable summary always tells the reader which decision the chosen centre, spread and display are intended to support.
What this chapter covers
- 01
Mean, median and mode
- 02
Range, variance and standard deviation
- 03
Coefficient of variation and z-scores
- 04
Quartiles, interquartile range and boxplots
- 05
Skewness and outlier diagnosis
- 06
Tables, histograms, time plots and scatterplots
Worked example · free
Compare suppliers with the same average
- 1Confirm that both samples share the same six-day mean.
- 1Compare the standard deviations in the original unit of days.
- 1Identify the one-day supplier as more predictable because observations cluster more tightly.
- 1Limit the conclusion to the observed evidence and recommend checking collection conditions and future variation.
Key terms
- Median
- The middle ordered observation, resistant to extreme tail values.
- Standard deviation
- The typical spread around the mean expressed in the variable's original unit.
- Interquartile range
- The distance from the first to the third quartile covering the middle half of observations.
- Standard score
- An observation's signed distance from the mean in standard-deviation units.
- Outlier fence
- A rule-based boundary that flags an observation for investigation rather than automatic deletion.
Descriptive Measures and Visualisation FAQ
When should I report the median?
Use the median when strong skew or extreme observations make the mean unrepresentative of a typical case. Pair it with the interquartile range so the centre and spread describe the same robust middle portion of the distribution.
What does a negative z-score mean?
A negative z-score places the observation below the stated mean because the numerator is observed value minus mean. Its magnitude gives the distance in standard-deviation units. It does not by itself label the value good, bad or erroneous.
Should every outlier be removed?
No. An outlier rule flags a case for checking. The observation may be a data-entry error, a rare valid customer, a system outage or a genuine market shock. The correct action depends on provenance and the business question.
Exam move
For every numerical dataset, write one sentence each about centre, spread and shape before calculating a long list of measures. Recompute a small variance by hand so the denominator and squared deviations remain meaningful. Practise z-scores with a direction prediction before substitution. Sketch left-skewed, symmetric and right-skewed distributions and place mean and median on each.
When using a boxplot fence, calculate the boundary, compare the observation and state an investigation action. Match each chart to a variable type and explain what pattern it is meant to reveal. Build paired summaries rather than a catalogue: mean with standard deviation for a reasonably symmetric distribution, or median with interquartile range when skew and outliers make robust measures more informative.
Use two tiny datasets with the same mean but visibly different spread and explain the operational consequence before calculating. Then alter one extreme value and observe which statistics move most. Keep sample and population notation separate by labelling the denominator before arithmetic. For visualisation, redraw a misleading chart with an honest scale and describe how the conclusion changes.
Practise choosing among bar chart, histogram, time plot, boxplot and scatterplot from the variable types and the decision, not from appearance. When a z-score is negative, state below the mean before discussing whether the result is desirable. Treat an outlier fence as a diagnostic trigger: verify entry, collection conditions and business meaning, and document any exclusion rule before seeing the result.
Finish each exercise with one sentence naming who uses the summary and why that centre-and-spread pair supports their decision.
Working through Descriptive Measures and Visualisation in BUSS1020? Sia is AskSia’s AI Statistics tutor — ask any BUSS1020 Descriptive Measures and Visualisation question and get a clear, step-by-step explanation grounded in how BUSS1020 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.