(I recommend you complete the exercises about self-paced reading in the "Reaction times" module before completing this task.)

You probably have already learned about one way we can study how people read: self-paced reading.

Think about how a self-paced reading experiment works. (If you don't remember, you can go back to the "Reaction times" module and do the self-paced reading garden path experiment again.) Then answer the following two questions:

1. Describe what a self-paced reading experiment is—how it works, and what it measures.
2. Can you think of any reasons why results from a self-paced reading experiment might not be realistic, or might not be a good measure of how people naturally read? e.g., are there any important differences between how natural reading works and how self-paced reading works?

Here are a couple potential problems with self-paced reading that I could think of:

1. Button-pressing is unnatural. In normal reading in real life, you don't have to press a button to read each next word. When you do self-paced reading, maybe the act of constantly pressing the button distracts you a little bit and changes the way you read.
2. No backtracking. When you read normally in real life, you can look back and re-read things (for example, maybe you read a few sentences and then realize you hadn't fully understood something from earlier, so you look back to read it again). But in the self-paced reading experiment we did, there's no way to go back; once you click the button and go to the next word, the previous words are gone.

These are two possibly important ways that self-paced reading is different than regular reading. There are probably other important differences, too; did you think of any other ones that I didn't mention here?

In any case, because of these issues, you might be skeptical about self-paced reading: do results from self-paced reading experiments really teach us about how normal reading happens?

One way to address that question is to use eye-tracking. Eye-tracking is a special method that uses some special software and a special camera to figure out where people are looking on a screen. In an eye-tracking experiment, you show the whole sentence on the screen (instead of showing one word at a time like you do in self-paced reading) and let a person read it. While the person reads it, a camera (usually just under the screen, or on a special hat that a person wears to hold a camera near their eye) records the person's eyeball. Using math (especially trigonometry—because the space between the screen, the camera, and the person's eye makes a triangle) the camera can figure out what place on the screen the person's eyes are looking at. Using this, the eye-tracker can figure out how long a person takes to read each word.

As I mentioned, there are two kinds of eye-tracking cameras: cameras that sit on the table usually below the screen (these are sometimes called "remote" eye-trackers), and cameras that a person wears on a special kind of hat (these are sometimes called "head-mounted" eye-trackers). Search online (using Google Images or something similar) and find a picture of what a remote eye-tracker looks like and a picture of what a head-mounted eye-tracker looks like.

To get a sense for how people's eyes move when they are reading, watch three brief videos on YouTube:

• Eye movements -- this is what the camera "sees" when it records someone's eyes while they are reading
• Eye-tracking demo #1 -- this shows where someone's eyes are looking while he is reading. The person reading just sees the screen with the text; he doesn't see the blue dot. The blue dot was added to this video later using data from the eye-tracker. The blue dot shows where the person was looking.
• Eye-tracking demo #2 -- this is similar to the demo #1 above, it shows where someone's eyes were looking while they were reading. This one shows it with a red line instead of a blue dot, but it's the same idea.

After having watched those, please describe how people's eyes move when they read. Some things you can address in your description:

• Do people's eyes move smoothly across the page, or is the movement kind of jumpy?
• Do people's eyes always move forward (from one word to the next word)?
• Do people's eyes stop to read every word?
• Is there anything that surprised you about the way people's eyes move (anything that you didn't expect, or that was different from what you previously thought)?