AP® Computer Science Principles · Big Idea 2 · Topic 2.4
Visualizing Data and Charts That Mislead
A chart can use completely accurate numbers and still mislead, because the way it is drawn, its axis range, its scale, its chart type, shapes what the viewer feels before they read a single number.
A bar chart tells the eye that height means amount, and the eye assumes the bottom of the bar is zero. Break that assumption by truncating the axis, starting it at 90 instead of 0, and you stretch tiny differences into towering cliffs. The math is simple: on a truncated axis running from s to 100, a bar's visible height is no longer its true value, it is how far that value sits above s, divided by the shrunken range from s to 100. Shrink the range enough and a 2 percent gap looks like one bar is twice the other.
Picking the right chart is its own skill, separate from axis tricks. Bar charts compare separate categories. Line charts show change over time or a trend. Pie charts show parts of one whole, and only work if the parts actually add up to that whole. Scatter plots show the relationship between two numeric variables. Using a line chart to compare unrelated categories, or a pie chart for data that does not sum to 100 percent, confuses a viewer even without any axis trickery at all.
This sits right after color and compression in the data unit because it is the last step: once bits become numbers, those numbers get drawn for a human to judge. The chart is an argument, not a neutral printout, and the same honest data set can be built to whisper or to scream depending on axis choice, scale, and chart type. Reading a chart well means checking the axis and the chart type before trusting the shape.
Key terms
- Truncated axis
- An axis that starts above zero, compressing the visible range and exaggerating differences between bars.
- Scale
- The mapping between a value and the space it takes up on a chart, which changes how big a difference looks.
- Chart type
- The style of chart, such as bar, line, pie, or scatter, chosen to match the kind of comparison the data supports.
- Misleading impression
- A false sense of the data's meaning created by how a chart is drawn, even when the underlying numbers are correct.
Watch it explained
- How it worksHow does starting a bar chart axis at 90 instead of 0 blow up a tiny difference?
- How it worksWalk me through picking a chart type that actually fits the question being asked.
- Worked exampleShow me a 2 percent change drawn on a truncated axis and then on a full axis.
- Common mistakeWhy isn't a chart automatically trustworthy just because the numbers behind it are real?
Check yourself
End of topic.