University of Sydney · FACULTY OF STATISTICS

QBUS5001 Chap.2 Descriptive Statistics and Association

- one subject, every graph, every model, every mark
6 Chapters4-page Bible
Our own words - no uploaded lecturer files
Updated for this semester
Chapter 2 of 10 · QBUS5001

Descriptive Statistics and Association

This chapter builds a dataset profile from measurement type, centre, spread, shape and unusual observations. It explains why mean, median and mode answer different questions and why variability must accompany a centre. The association section distinguishes covariance, correlation and causation, with a scatterplot check for curvature, clusters and influential observations before any coefficient becomes a business conclusion.

Measurement level controls which operations make sense: a code can label a category without carrying numerical distance, while quantitative values support differences and ratios only when their scale permits them. The mean uses every magnitude and therefore moves toward an extreme observation; the median depends on order and is more resistant; the mode identifies frequency rather than balance.

Range, interquartile range, variance and standard deviation describe different aspects of spread, so a centre without variability hides how dispersed the business outcomes are. Covariance records joint direction but retains the units of both variables. Correlation standardises that linear association, yet a large value can still conceal curvature, group structure or a single influential point.

The data profile and scatterplot therefore come before the coefficient, not after it.

In this chapter

What this chapter covers

  • 01

    Population parameters and sample statistics

  • 02

    Nominal, ordinal, interval and ratio measurement

  • 03

    Mean, median and mode under skew and outliers

  • 04

    Range, interquartile range, variance and standard deviation

  • 05

    Covariance direction and measurement units

  • 06

    Correlation strength, scatterplots and causal boundaries

Worked example · free

Summarise a skewed delivery sample

Q [5 marks]. Delivery times are 18, 20, 21, 22, 24, 25 and 70 minutes. Choose a centre and spread for typical service and explain the extreme value. The mark allocation used here is not a University assessment scheme.
Centre and spread respond differently to an outlierTwo dot plots show the same six values before and after the largest value moves far right. AB centre median stable mean pulled right
The lower dot plot moves only the largest observation, leaving the median near the centre while pulling the mean right and increasing range and variance.
  • +1Order the data and identify the median as 22 minutes.
  • +1Compute the mean as approximately 28.57 minutes and explain its sensitivity to 70.
  • +1Use the interquartile range of 5 minutes under the median-of-halves convention for the middle spread.
  • +1Investigate 70 as a possible process failure or data error rather than deleting it automatically.
  • +1Report median and IQR for typical service while preserving the late case as operational risk evidence.
Median 22 minutes and IQR 5 minutes describe typical service better than the mean. The 70-minute case should be investigated and reported separately.
Sia tip — A robust summary protects the typical value from an outlier; it must not hide the outlier's operational consequence.
Glossary

Key terms

Population Parameter
A fixed numerical feature of the population of interest.
Sample Statistic
A numerical feature calculated from a sample and used to estimate a parameter.
Interquartile Range
The distance from the first quartile to the third quartile, covering the middle half of ordered data.
Sample Correlation
A unitless measure of the direction and strength of linear association in sample data.
FAQ

Descriptive Statistics and Association FAQ

Should I report the mean or median?

Choose the statistic from the question and distribution. The mean uses every magnitude and suits totals or balanced data; the median resists extreme observations and often describes a typical value better in a skewed distribution. Reporting both can reveal the skew.

Why is standard deviation not enough on its own?

Standard deviation describes spread around the mean but does not show skew, clusters or unusual observations. Pair it with a centre, sample size and distribution plot so the reader can see which pattern generated the summary.

Does high correlation prove causation?

No. High correlation describes a strong linear association in the observed data. Confounding variables, common trends, selection or influential observations may create it. A causal claim requires a design and assumptions that isolate the effect of changing the predictor.

Study strategy

Exam move

For every dataset, write a five-line profile: measurement type, centre, spread, shape and unusual observations. Recompute one example after adding an outlier to see which summaries move. For two variables, sketch a scatterplot before calculating correlation, then describe direction, form, strength and influential points. End by separating observed association from a causal mechanism.

Create paired examples that force a choice rather than a calculation. Give two distributions the same mean but different spread, then explain which business risk is hidden by reporting only the mean. Give a skewed sample a mean and median, then connect the preferred centre to the question being asked rather than calling one statistic universally better.

For association, preserve a short checklist beside every scatterplot: variable roles, units, direction, form, strength, clusters and influential observations. Recalculate covariance after changing measurement units to see why its magnitude is not directly comparable, and verify that correlation stays unchanged. Finally, write a causal sentence and edit it down to the strongest association sentence supported by the design.

That editing exercise is as important as computing the coefficient because interpretation carries the decision meaning. Keep calculations and descriptions linked by annotating each statistic with the feature it captures and the feature it misses. A mean summarises balance but not skewness; a standard deviation summarises distance from the mean but not multimodality.

This habit stops a compact summary from replacing inspection and makes the final business explanation more faithful to the observed distribution.

Working through Descriptive Statistics and Association in QBUS5001? Sia is AskSia’s AI Statistics tutor — ask any QBUS5001 Descriptive Statistics and Association question and get a clear, step-by-step explanation grounded in how QBUS5001 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.

A+Everything unlocked
Unlocks this Bible + all 202 of your subjects at University of Sydney, plus 1,000+ more Bibles across the AskSia library.
Sia - your QBUS5001 tutor, unlimited, worked the way the exam marks it
The full 4-page Bible + practice bank with worked solutions
Chrome extension - sync your LMS so Sia knows your deadlines
Bilingual EN / Chinese on every Bible and every Sia answer
$0.99 Trial
30-day money-back · cancel in one tap · how it works
QBUS5001 · Foundation in Data Analytics for Business - independent study guide on the AskSia Library. More University of Sydney subjects · Microeconomics across all universities