Data visualization can be very powerful when done well. Your visual cortex—the part of your brain that processes visual images—is the largest system in your brain, and it’s great at spotting patterns. When you visualize data, you can use your visual perception to quickly identify patterns in data and more easily pick out important patterns that you wouldn’t see in a table or list of numbers.

We see graphs a lot but often don’t know how to approach them. Here are the steps we’d recommend, using this graph as an example:

### 1. Examine the component parts.

Read the **title**. “This Month’s vs Last Month’s Algebra Test Scores”. Okay, we’re comparing two months of Algebra scores.

Check out the **axis labels**. The x-axis is labeled “Last Month’s Score”; the y-axis “This Month’s Score”. So this is where the comparison happens: one score is on the vertical scale, and the other is on the horizontal. Those are the reverse of what we’d usually expect to see, especially given the chart title. Good thing we looked!

Look at the **axis scales**. In this case, both axes range from 40 to 100, so we’re dealing with roughly the upper half of the test range in both cases.

### 2. Think with your eyes.

Use your visual cortex. What do you see here? Does any distinct pattern stand out? Here are a few thoughts:

- The points generally follow a diagonal path, or trend, from the lower-left-hand corner to the upper-right-hand corner of the graph. So students seem to do roughly as well on the second test as they do on the first test.
- There are a bunch of points pressed up against the upper-right-hand corner of the graph. Why is that? The test maxes out at 100, but some students would do even better if the test could register a higher score. In other words, we may want to consider making future tests harder so that we can see where all students fall on the range.
- Several individual points (representing individual students) lie outside of the rest of the herd. These are called outliers, and we will investigate them more in a later lesson.

### 3. Zoom out for the big picture.

To get a big picture understanding of a graph, ask yourself these questions:

*What is the author of the graph trying to communicate?*In this case, the author seems interested in comparing how much students improved from one month to the next.*How is this data relevant to my classroom?*If it’s a graph of your students, it may help you identify that some students are exhibiting the most rapid growth, and others are struggling.*Is the graph biased?*Does the author have any personal stake in presenting this data a certain way? Does the data reflect reality? Is it presented in an objective way?

You can remember these steps as PEZ—Parts, Eyes, Zoom—like the candy dispenser, because being able to quickly digest graphs and to recognize trends is suh-weet! And with a bit of practice, you’ll be able to carry out all of these steps in a matter of a few seconds.

If you’re looking for a far more comprehensive guide to reading and creating graphs, you’ll do no better than the fancy-sounding but easy-reading Now You See It: Simple Visualization Techniques for Quantitative Analysis by information design guru Stephen Few.

## Exercise: Reading a graph

Check out the following graph from the National Association for Education Progress. See if you can decipher what it is trying to tell you.

## #Takeaways

- Graphs leverage the pattern-matching power of your visual cortex. Tweet

- Save pie charts for dessert; use bar charts for data. Tweet

- I just learned how to think with my eyes! Tweet

We’ve now learned three of the most common skills you’ll need to work with data: Data collection, summary, and visualization. Everything we do going forward will build on these basics. In the next lesson, we’ll head back to our trusty spreadsheet to learn how to create good-looking graphs in just a couple clicks.

I agree with the statement, “save the pie for desserts”. I think if you’re serious about looking at data, either plot it or use a bar graph. This allows you to see growth, degression, patterns, etc.

I like using a bar graph to collect data.

Graphs help make sense out of numbers.