2.6. Misleading Graphs

Learning Outcome

Students will be able to locate data visualizations and deconstruct the graph in order to evaluate the effectiveness of the visualization.

Sample Tasks

• Find graphics in the “wild” and bring them into class to critique.

  • Example of bad graphs could be the following:

    • Non-zero axis

    • Zero axis

    • Pie charts versus bar charts versus histograms

    • Inappropriate pictograms

  • The goal of the project is for students to find intentionally misleading graphs.

  • Students will critique each example of a bad graph and explain why it is inappropriate in the given situation.

  • Students will propose a more appropriate presentation for the same data and explain why that presentation is the better choice.

  • Students will proceed to create an example using the appropriate presentation.

  • In the end the students present the inappropriate and appropriate graphs to the class for side-by-side comparison.

• Choosing to highlight proper information.

  • Student will choose variables they want to highlight from a larger set of variables.

  • Student will need to decide what they want to show.

    • Cases

    • Hospitalizations

    • Deaths

  • Student will explain the importance of one set of variables versus another.

    • Student will experiment with different color-coding schemes to highlight important variables, and decide between the following:

      • Standard pallet

      • Custom pallet

    • Students will justify their choice.

    • Student will discuss disability issues such as color blindness.

[OhioDoHEducation21]

A visualization can be misleading if either it is poorly done or if the creator is trying to deceive.

Our first readings, from Learning Data Science [LGN23], are repeated from Section 2.3. These remind us how to plot well, to avoid unintentionally being misleading.

• 11. Data Visualization

  • 11.1. Choosing Scale to Reveal Structure

  • 11.2. Smoothing and Aggregating Data

  • 11.3. Facilitating Meaningful Comparisons

  • 11.4. Incorporating the Data Design

  • 11.5. Adding Context

Our second reading, from Computational and Inferential Thinking [ADW21], is repeated from Section 2.2. It describes a principle that is often violated in advertising to distort relative sizes.

• 7.2.4. The Area Principle

Our third reading, from Wikipedia, describes methods used to create misleading graphs.

• Misleading graph

Reading Question

• What is the Lie factor?

Our fourth reading, from Statistics How To, discusses several examples of misleading graphs used in the press.

• Misleading Graphs: Real Life Examples

Reading Questions

• Which misleading graph angered you the most?

• How long would it take you to find a misleading graph on the internet?