Choosing Appropriate Graphs to Display Data
Learning Objective
- Choose appropriate graphs and charts to display data
Why This Matters
Every election cycle, news outlets publish graphs that make a 2-point polling shift look like a landslide or flatten a genuine trend into a straight line -- same data, different graph choices, completely different stories. The same thing happens in corporate earnings reports, public health dashboards, and every viral chart on social media. Choosing the right graph isn't a design preference -- it's the difference between informing your audience and misleading them.
How to Use This Simulation
- Read each scenario card describing a dataset and what the presenter wants to show.
- Select the best graph type from the four options. Use the Decision Guide on the right for help.
- Review the feedback -- you'll see the correct graph alongside the misleading alternative to understand why the choice matters.
- Check the Explanation Panel below -- it updates after each card and builds toward a decision framework you can use on any dataset.
Graph Misuse Patterns You Encountered
Decision Guide
What's Happening
Quick Check
A local news website publishes a line graph showing average home prices in five neighborhoods: Riverside ($320K), Downtown ($410K), Oakdale ($285K), Lakewood ($395K), and Hillcrest ($305K). The line slopes upward from left to right, and the caption reads "Home prices rising across the city." A reader shares the graph with the comment "Finally, the housing market is recovering!" What is the fundamental problem with this graph?
Try This
A campus bookstore tracked the prices of 28 required textbooks this semester. The manager wants a single display showing students the overall price distribution so they can see where most textbook prices cluster. Which graph type should the manager use? Use the Decision Guide's logic to work through the answer, then explain in one sentence why your choice best serves this scenario.
A fitness tracker app wants to display a user's total calories burned for each of the past 14 days. Both a bar graph and a line graph could display this data correctly. (1) Name both reasonable options and what each communicates. (2) Recommend one, explaining which question it answers best. (3) In one sentence, explain what the rejected option would communicate less effectively.
A news website publishes a bar graph showing unemployment rates for four cities. The y-axis starts at 4.0% instead of 0%, making City A's rate of 5.2% look roughly triple City D's rate of 4.4%. The headline reads "Unemployment Gap Widens Between Cities."
(1) What graph type was used? (2) What does the graph misrepresent about the actual differences? (3) Recommend a y-axis treatment that would communicate the data more honestly. (4) In two sentences, defend your recommendation addressing both technical accuracy and whether the original graph's visual proportions match the data's actual story.
Instructor Notes
Teaching Notes
This simulation works best when you let students encounter Card 5 (Average Rent by Neighborhood) without warning. Most students will have correctly classified Cards 1-4 and built confidence. Card 5's data is numerical (dollar amounts), which tempts students toward quantitative graph types -- but the x-axis variable is qualitative (neighborhood names). The line graph rendering with the "Shuffle Neighborhoods" button is the centerpiece: students see the same data produce opposite "trends" depending on category order.
The Decision Guide panel is designed to be a scaffolding tool, not a crutch. After the simulation, ask students to close the guide and classify a novel scenario from memory. The guide's three-question framework (variable type, natural order, sample size) should become internalized.
Common Student Errors
- Choosing line graphs for categorical data because the data values are numerical. The variable type on the x-axis, not the y-axis, determines the graph choice.
- Defaulting to histograms for all quantitative data regardless of sample size. With n under 50, stem-and-leaf plots preserve individual values that histograms aggregate away.
- Treating graph choice as purely aesthetic ("I just like how bar graphs look") rather than methodological. The simulation is designed to surface consequences of wrong choices.
- Assuming that a graph must be mathematically incorrect to be misleading. Truncated y-axes and line graphs on categories are technically plottable but visually dishonest.
Discussion Questions
- Find a graph in a recent news article or social media post. What graph type was used? Was it the right choice for that data? Could a different choice have told a more honest story?
- A company's quarterly report uses a bar graph with a y-axis starting at $4.8 billion instead of $0 to show revenue growth from $5.0B to $5.3B. Is this misleading? When, if ever, is it acceptable to start the y-axis above zero?
- If you had to explain the Decision Guide to someone who missed this lesson, could you do it in three sentences?
Exam Connection
Typical exam questions present a data description and ask students to identify the appropriate graph type. The Decision Guide's three-question framework (variable type, order, sample size) maps directly to how these questions are structured. Card 7 (the judgment-call card) specifically prepares students for questions where more than one answer could be defensible -- exams occasionally include "which is the best choice" items where students must justify their reasoning.