A chart can make a pattern obvious in one second — or make a tiny difference look enormous using nothing but true numbers. Today: which chart for which question, and how to catch the lie.
No reading — this page is the reading. Your teacher hands out the Lab 5.8 spec today; the chart builder below is how you’ll check your thinking before Part 4.
Big Data. You can clean, filter, and group. Now: how to show an answer so a person gets it in one look — and how a chart can mislead without containing a single false number.
Same twelve numbers. One way took you several seconds of reading; the other took one glance. That’s what a visualization is for.
A visualization turns numbers into a picture. The question decides the chart:
| If the question is about… | Use | Example |
|---|---|---|
| comparing amounts across categories | bar chart | members per club |
| change over time | line chart | checkouts per month |
| a relationship between two numeric variables | scatter plot | practice hours vs free-throw % |
Every chart has axes, and every axis has a scale — where it starts, where it ends, and how far apart the marks are. The scale is where charts lie.
Pick a dataset, then a chart type. Read the note: does the chart fit the question? Then flip Mislead me and watch the same numbers tell a different story.
Clean the club data (Part 1), show that cleaning changes the answer (Part 2), group by grade and make a bar chart (Part 3), then make the same numbers lie (Part 4). A spreadsheet is fine for the charts — the choice and the axis are the point, not the tool.
Five questions. Name the question first; the chart follows.
That’s what makes it effective. Nobody can call it a lie — every value is correct. Ads, campaign flyers, and news graphics use truncated axes, cherry-picked date ranges, and stretched scales because they work. The defense is a habit: read the axis before you read the bars.