Chapter 3 — Graph and the Chart-Pencil Set

Graph was a small finch. She was also a tween. She had bright yellow feathers. They mixed with cream and warm russet colors. Her eyes sparkled. She always carried a tiny leather case. It held her chart-pencils. She kept it tucked into her vest pocket.

Graph moved quickly. She was very careful about colors. Every color had to be just right. Her chart-pencil case held eight pencils. Each one was a different color. It also held a small, folded card. This card listed different chart types. It said: bar / line / scatter / pie / histogram / box-plot / heatmap / map.

When Graph saw new data, she got to work. First, she unfolded her chart-type card. She looked at the data’s shape. Then she picked the right tool from her list. Finally, she drew the chart. She always chose a color that fit the data’s mood.

This was a very important job. Graph taught everyone about data-visualization. This meant choosing a chart. The chart had to show the data honestly. It was a key skill for working with data.

Lots of new students made the same mistake. They picked charts because they looked cool. Or because they filled up a lot of space. Sometimes, they just used all the colors. Graph said this was wrong. A chart should not be just pretty.

The best chart was the one that matched the data’s shape. Bar charts helped compare different things. Line charts showed how things changed over time. Scatter plots showed how two things were connected. Pie charts showed parts of a whole. But they only worked if the parts added up to 100%. Histograms showed how one thing was spread out.

Every chart type had a special job. The data itself had a shape. That shape always matched one (or maybe two) of those jobs. Graph knew this well.

Graph never thought of charts as just decoration. She was very clear about it. “Charts are not just pretty pictures,” she would say. “A chart is an argument. Every chart you pick says something about the data.” She would tap her pencil. “The chart you choose changes how people see the data. The chart that tells the truth is not always the loudest one.”

This was important. Many popular charts looked fancy. They had 3D pie charts. They had silly animations. They used rainbow colors. But these often hid the data. They made it harder to see the truth. Graph taught that a chart should show the truth. It was not just for grabbing attention.

She also taught about charts that lied. She showed students how charts could trick people. Things like cutting off the bottom of a chart. Or using two different scales that didn’t match. Or making charts look 3D when they shouldn’t. She taught about picking only certain times. Or missing starting points. Kids learned to spot these tricks. They learned to avoid them in their own work. The real skill was seeing past the chart. It was about seeing the data underneath.

Graph grew up in a small village. Her family made quilts for the village. They were the finches who designed the yearly harvest-quilt. Each square on the quilt showed what a family had given. The colors were chosen to honor the data. This work taught Graph a lot. She learned that colors and shapes carried meaning. A quilt with flashy, random colors was not trusted. But a quilt with meaningful colors became a family treasure. By age six, Graph knew something deep. Visualization was about being honest. The chart that earned trust was the chart that honored the data.

When she was twenty-two, Graph walked to the DataForge academy. Datum, the head of the academy, asked her a question. “What is data visualization?” Datum asked.

Graph thought for a moment. She looked at Datum. “It is the shape of the story,” she said. “The chart that tells the truth is not always the loudest one.” She paused. “The chart you choose says something about the data. You must match the chart to the data’s shape. And you must teach people to see through the chart. They need to see the data itself.”

Datum smiled. “You are appointed,” Datum said.

In her workshop, Graph started every first lesson the same way. She carefully unfolded her chart-type-reference card. The paper crinkled softly. Then she opened her small leather chart-pencil case. The pencils lay neatly inside.

“I am Graph,” she would say. Her voice was clear. “The skill I teach is visualization. The main idea is this: match the chart to the shape of the data.” She pointed to her card. “Bar charts are for categories. Line charts are for time. Scatter plots are for relationships. Pie charts are for parts of a whole. Histograms are for distributions. Each chart has a job. Pick the chart that fits.”

She taught her students how to build good charts. These were her steps:

• Find the data’s shape. Is it about groups? Is it always changing? Does it show time? Is it about places? How is it spread out? Does it show connections?
• Match the chart to the shape. Groups? Use a bar chart. Changing over time? Use a line chart. Two changing things? Use a scatter plot. Parts of a whole that add to 100%? Use a pie chart. How one thing is spread out? Use a histogram. About places? Use a map.
• Don’t pick the flashy one. 3D charts usually show data worse than flat 2D charts. Rainbow colors are usually worse than colors you picked for a reason. Pie charts with more than five parts are usually worse than bar charts.
• Start the bottom line at zero (most times). If you cut off the bottom of a chart, small differences look huge. Sometimes it’s okay, like with temperature. But you must label it clearly.
• Label everything. Label the sides of the chart. Label the units. Show the time range. Say where the data came from. Tell how many things were counted. The chart should tell the viewer everything they need to know.
• Test the chart. Ask yourself: “What does this chart claim?” If the data doesn’t really support that claim, the chart is lying. Make a new one.
• Learn the lying charts. Kids need to spot these in the real world. Cut-off charts. Charts with two different scales that don’t match. Squished 3D charts. Charts that only show a small, good part of the time. Charts missing their starting point. Charts that make small areas look like big volumes.
• The chart is an argument. Charts are not just neutral pictures. Every chart choice says something. Make sure your claim is honest.

Graph was very honest herself. “Sometimes I draw a chart that looks beautiful,” she would say. “But it hides the data. That’s okay. That’s how I learn. I learn when pretty things and honest things pull apart.” She would tap her pencil again. “Making it better is the practice. Honesty first, then beauty.”

When students asked Graph if visualization was hard, she always gave the same answer.

“It is not hard,” she would say. “It is just match the chart to the shape of the data. The chart that tells the truth is not always the loudest one.”

Then she would close her chart-pencil case. The next dataset waited. It was ready to be charted.

---