Learning Objectives

By the end of this lesson, learners should be able to:

  • Explain data distributions.
  • Identify symmetric and skewed distributions.
  • Detect outliers.
  • Interpret business implications of skewness and outliers.

Understanding Data Distribution

A distribution describes how values are spread across a dataset.

Symmetric Distribution

Values are balanced around the center. Mean, median, and mode are approximately equal.

Positively Skewed Distribution

Long tail to the right. Example: executive salaries or luxury purchases.

Negatively Skewed Distribution

Long tail to the left. Example: customer satisfaction scores when most customers are highly satisfied.

Understanding distribution helps analysts choose appropriate statistical measures.

Histograms

Histograms display the distribution of continuous data.

Example

A courier company analyzes delivery times and finds that most deliveries occur within two days, while a small number take much longer. The histogram reveals a positively skewed distribution.

Managerial Action

Investigate delayed shipments, customs issues, or transportation bottlenecks.

Box Plots

Box plots summarize:

  • Median,
  • Quartiles,
  • Spread,
  • Potential outliers.

They are particularly useful for comparing regions or business units.

What Is An Outlier?

An outlier is an observation that differs substantially from the rest of the data.

Possible Causes

  • Data entry errors,
  • Fraud,
  • Exceptional business events,
  • Operational failures,
  • Measurement errors.

Outliers should be investigated rather than automatically removed.

Interquartile Range (IQR) Method

  • IQR = Q3 − Q1
  • Lower bound = Q1 − 1.5 × IQR
  • Upper bound = Q3 + 1.5 × IQR

Values outside these bounds are potential outliers.

Worked Example

Suppose:

  • Q1 = 40
  • Q3 = 60

IQR = 20

Upper bound = 60 + 1.5×20 = 90

Any value above 90 is a potential outlier.

Business Case: Fraud Detection

A payment processing company identifies several unusually large transactions that exceed the IQR upper bound. Analysts investigate and discover fraudulent activity.

Impact

Early detection prevents financial loss and improves compliance.

Skewness In Business

Customer Spending

A small number of customers often account for a large share of revenue.

Property Prices

Luxury properties create positive skewness.

Insurance Claims

Most claims are small, but a few catastrophic claims create extreme right tails.

Managers must recognize skewness when setting budgets, forecasts, and performance targets.

Common Mistakes

  • Deleting outliers without investigation.
  • Assuming all skewed data is problematic.
  • Using mean alone for heavily skewed data.
  • Ignoring business context.

Practical Exercise

Create a histogram and box plot for customer spending data from five international markets and identify potential outliers.

Learning Materials / Reference Materials

  • OpenIntro Statistics.
  • NIST Distribution Analysis Resources.
  • Tableau Histogram and Box Plot Guides.

Lesson Summary

Understanding distribution and outliers helps analysts detect unusual patterns, operational problems, fraud risks, and important business opportunities.

PreviousNext