Financial Management and Business Data Analytics · Data Presentation: Visualisation and Graphical Presentation

Scatter Plots, Heat Maps and Advanced Charts

Updated 10 October 2026 · Fact-checked

A scatter plot shows the relationship between two numeric variables as dots. A bubble chart adds a third variable as dot size. A heat map uses colour to show values in a grid. A box plot shows spread and outliers. A treemap shows parts of a whole as nested rectangles. Pick the chart that matches your question.

Understand Scatter Plots, Heat Maps and Advanced Charts

Bar charts and line graphs show one measure at a time. Many business questions ask something else: do two things move together, where are the hotspots, how spread out is the data, or which part is biggest? The charts in this topic answer those questions.

A scatter plot (scatter diagram) puts one numeric variable on the X axis and another on the Y axis. Each observation is one dot. If dots rise from left to right, the relationship is positive. If they fall, it is negative. If they look like a shapeless cloud, there is little or no linear relationship. A tight band means a strong relationship; a wide band means a weak one. Dots far from the pattern are outliers. A scatter plot shows association, not cause. Advertising spend and sales may rise together because of a festival season.

A bubble chart is a scatter plot where the size of each dot shows a third numeric variable. For example, X is market share, Y is profit margin and bubble size is turnover. A heat map is a grid (rows by columns) where colour intensity shows the value in each cell. It is good for spotting patterns in many numbers at once, such as sales by region and month, or a correlation matrix. Always include a colour legend.

A box plot (box-and-whisker plot) summarises a distribution using five numbers: minimum, first quartile (Q1), median, third quartile (Q3) and maximum. The box runs from Q1 to Q3 and a line inside marks the median. Whiskers extend to the smallest and largest values that are not treated as outliers, and outliers are shown as separate points. It lets you compare spread across groups, for example branch-wise delivery times. A treemap divides a rectangle into nested smaller rectangles whose areas are proportional to values. It shows the share of each category in a total, and works well when there are many categories, such as expenses by department.

Key rules to remember

Interquartile range (IQR): IQR = Q3 − Q1; Length of the box in a box plot. It holds the middle 50% of the data.
Outlier fences (common 1.5 × IQR rule): Lower fence = Q1 − 1.5 × IQR; Upper fence = Q3 + 1.5 × IQR; Values outside the fences are usually plotted as outliers. This is a convention used by many tools, not a law.
Treemap area: Area of a rectangle ∝ category value ÷ total of all values; Bigger area means bigger share.
Reading a scatter plot: Upward trend = positive; downward trend = negative; no pattern = no linear relationship; Association does not prove causation.

How to solve Scatter Plots, Heat Maps and Advanced Charts questions

For any question on choosing or reading an advanced chart, use this method.

1Identify the business question: relationship, pattern across a grid, spread, or share of a total.
2Count the variables and note their type (numeric or category). Two numeric variables point to a scatter plot; three numeric point to a bubble chart.
3Match the chart: scatter or bubble for relationships, heat map for grid patterns, box plot for spread and outliers, treemap for parts of a whole.
4If reading a chart, state the direction, strength and any outliers or clusters.
5For box plots, compute IQR and the fences if values are given, and list outliers.
6Write the business meaning in one line, and add a caution that association is not causation.
7Mention a limitation or an alternative chart if the question asks for justification.

Quickest way: Question-to-chart shortcut

When to use it: Use for MCQs and short 'which chart' questions.

Two numbers, relationship: scatter plot.
Add a third number as size: bubble chart.
Colour-coded grid of values: heat map.
Spread, median, outliers by group: box plot.
Share of a total, many categories: treemap.
Eliminate options that show only one variable or only trends over time.

Common mistakes in Scatter Plots, Heat Maps and Advanced Charts

Saying a scatter plot proves that X causes Y.
A clear upward pattern looks like cause and effect.
Fix: Say the variables are associated. Cause needs separate evidence.
Using a scatter plot for a category against a number.
Students forget scatter plots need two numeric axes.
Fix: Use a bar chart or box plot when one variable is a category.
Reading the box plot whiskers as minimum and maximum always.
Textbook diagrams often show whiskers reaching the extremes.
Fix: When outliers are plotted, whiskers stop at the last non-outlier value.
Treating the line in the box as the mean.
Mean and median are confused.
Fix: The line in a box plot is the median.
Using a treemap or heat map without a legend or labels.
Students assume colour and size are obvious.
Fix: State what colour and area represent, and label units.
Choosing a bubble chart when size adds nothing.
Students like complex charts.
Fix: Use a bubble chart only when a real third numeric variable exists.

Worked examples

Example 1

A firm records monthly advertising spend (₹ lakh) and sales (₹ lakh) for 12 months. Dots on the plot rise steadily from left to right in a narrow band, with one dot far below the band. Which chart is this, and how do you interpret it?

Show the solution

Two numeric variables are plotted as dots, so the chart is a scatter plot.
The dots rise left to right, so the relationship is positive.
The band is narrow, so the relationship is strong.
The dot far below the band is an outlier. Check that month for a data error or special event, such as a stock shortage.
Caution: the plot shows association. Other factors may also affect sales.

Answer: A scatter plot showing a strong positive association between advertising and sales, with one outlier to investigate. It does not prove advertising causes the sales.

Example 2

Delivery times (days) for a branch are: 2, 3, 4, 5, 6, 7, 8, 9, 10, 25. Q1 = 4, median = 6.5, Q3 = 9. Find the IQR, the outlier fences and any outliers for the box plot.

Show the solution

IQR = Q3 − Q1 = 9 − 4 = 5.
Lower fence = 4 − 1.5 × 5 = 4 − 7.5 = −3.5.
Upper fence = 9 + 1.5 × 5 = 9 + 7.5 = 16.5.
No value is below −3.5.
The value 25 is above 16.5, so it is an outlier.
The upper whisker stops at 10, the largest value within the fences.

Answer: IQR = 5 days; fences are −3.5 and 16.5; the only outlier is 25 days. The box spans 4 to 9 with the median line at 6.5.

Exam tips

For 'which chart' MCQs, count the variables first. This removes two or three options quickly.
Learn the five numbers of a box plot and the IQR fence rule. Numerical questions often ask for them.
In written answers, always state direction, strength and outliers for a scatter plot, then the business meaning.
Give one business example for each chart. Examples earn step marks.
Write the limitation 'association is not causation' in any interpretation answer.

Practice questions from Data Presentation: Visualisation and Graphical Presentation

Scatter Plots, Heat Maps and Advanced Charts: frequently asked questions

What is a scatter diagram and how do I interpret it?

It plots two numeric variables as dots, one dot per observation. Look at direction (up is positive, down is negative), strength (how tight the dots are) and outliers. It shows association, not cause.

When should I use a bubble chart instead of a scatter plot?

Use a bubble chart when you have a third numeric variable to show as dot size. If you have only two variables, a scatter plot is cleaner.

What does a box plot show?

It shows the minimum, Q1, median, Q3 and maximum, so you see the centre, spread and outliers. It is useful for comparing several groups side by side.

When is a treemap better than a pie chart?

A treemap handles many categories and nested groups better. A pie chart becomes hard to read with many slices.