24761 Chap.5 Choosing the Right Data Visualisation
Choosing the Right Data Visualisation
Why charts work
The Week 4 lecture names two routes for processing information, systematic and heuristic, and argues that charts help because visuals are easier to understand than numbers or text, simple to share and quicker to draw conclusions from.
Three questions, four purposes
Before choosing a chart, ask what type of data you hold, how many variables must be shown and what one message the reader should take away.
The lecture then sorts charts by purpose: comparison, relationship, composition and distribution.
Reading beyond the basics
The chapter audits a festival analyst’s four draft charts, draws a spend histogram to read its centre and tail, and covers the lecture’s other charts, from word clouds and radar charts to bubble charts and perceptual maps, with the mistakes each invites.
These choices feed both the first report, where you read charts inside sources, and the poster, where you build your own. Each chart choice is justified from the data, never from appearance.
What this chapter covers
- 01
Systematic and heuristic processing of visual information
- 02
Three questions to ask before choosing a chart
- 03
Comparison, relationship, composition and distribution
- 04
Reading a histogram: mode, median and tail
- 05
Word clouds, radar charts, bubble charts and perceptual maps
- 06
When a chart choice distorts the message
Worked example · free
Read the festival spend histogram
- 1The tallest bar is $25 to $50 with 380 attendees, so that is the modal band.
- 1Counting up, 140 attendees sit in the first band and 140 + 380 = 520 by the end of the second, so the 500th and 501st attendees, and therefore the median, fall in the $25 to $50 band.
- 1Counts fall away to 290, 120 and 70 as spend rises, so the distribution has a long right tail: most people spend modestly and a small group spends heavily.
Key terms
- Bar Chart
- Separate bars whose lengths compare a value across categories such as brands, segments or channels.
- Histogram
- Touching bars that count how many observations fall into each band of a numerical variable, revealing its shape and skew.
- Box Plot
- Built from five statistics, the minimum, quartiles, median and maximum, it compares spread and flags outliers across groups.
- Bubble Chart
- Essentially a scatter plot in which the size of each bubble carries a third variable.
- Radar Chart
- Spokes radiating from a centre, each one a measure on the same scale, so profiles across several attributes can be compared.
Choosing the Right Data Visualisation FAQ
How do I choose the right chart for marketing data?
Start with three questions from the Week 4 lecture: what type of data you have, how many variables you need to show, and what you want the reader to understand. Then match the purpose: bars or lines to compare, scatter or bubble charts for relationships, pies for parts of a whole, histograms or box plots for spread.
When should I use a pie chart?
Only when the slices are parts of one whole that adds to 100%, such as the share of tickets sold by tier. If the values do not add to a total, or there are many small categories, a bar chart sorted from largest to smallest is easier to read and harder to misjudge.
What is the difference between a histogram and a bar chart?
A bar chart compares separate categories, so its bars stand apart and could be reordered. A histogram groups a continuous measure such as spend or time into consecutive bands, so the bars touch and their order is fixed; its job is to show how values are distributed.
Does a big word in a word cloud mean customers liked it?
No. Size reflects how often a word appears in the text, not the feeling behind it, so a large word can come from mostly negative comments. Pair a word cloud with a reading of the comments or a sentiment count before you draw any conclusion about attitudes.
Which chart shows whether two things are related?
A scatter plot, with one numerical variable on each axis and one point per case. If you need a third variable, a bubble chart adds it as bubble size. Either chart can show that measures move together, but neither proves that one causes the other.
Assessment move
Take three charts from an industry report or news article you might cite and, for each, write the data type, the number of variables and the purpose it serves. Decide whether the chart matches that purpose and redraw any that do not. Then practise building the four core types yourself in Excel or a similar tool from a small invented dataset, because the lecture argues that making charts is what teaches you to read them.
Finish by writing one-sentence chart titles that state the point, a habit you will need for the poster. Keep the redrawn charts; they are rehearsal for the poster. Review your notes weekly.
Working through Choosing the Right Data Visualisation in 24761? Sia is AskSia’s AI Marketing tutor — ask any 24761 Choosing the Right Data Visualisation question and get a clear, step-by-step explanation grounded in how 24761 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.