Eye movements/eye-contingent experimental paradigms
Eye-Contingent Experimental Paradigms
This is a brief description of the three of the most ground-breaking paradigms used to discover just how much you “see” while you are reading.
- All of these are called “gaze contingent” because what a person sees is dependent on where they are looking. So, in order to conduct an experiment with any one of these paradigms, a highly-accurate eye tracker is needed.
Moving Window Paradigm
- The first is a Moving Window paradigm, in which in this example only the word being currently fixated is visible. It is possible to manipulate how many words or even how many characters is visible to the reader.
- The eye movement measures while reading with a window (at various sizes) are compared to eye movement measures while reading normally (i.e., no window).
- If the information in the parafoveal and periphery are being used during normal reading, then having a window that covers that information will slow down reading.
- To find where the limits are, we simply look for the smallest window size that does not produce any interference (i.e., the eye movement measures look the same as normal reading).
- To get an idea of what reading with a window is like Figure 1 shows a time sequence of a 1-word window.
- Figure 1: This figure shows what a one-word moving window looks like as a person is reading over time. In each frame, the word at the person’s current fixation is shown, while the rest of the sentence is blocked from view. Research has shown that reading with only a 1-word window is extremely difficult and slow.
Moving Mask Paradigm
- The second method used is the Moving Mask paradigm, which is essentially the inverse of the window and in this case, blocks out the current word being fixated.
- This paradigm is used to find out how important the central information is compared to the parafoveal and peripheral information.
- Researchers have found that reading with a mask is detrimental, and reading slows down dramatically while regressive saccades increase dramatically. It seems that without the central information, reading becomes nearly impossible.
- Figure 2 shows a time sequence of reading with a 1-word mask.
- This figure shows what a 1-word mask looks like as a person reads a sentence. The word at the current fixation is masked, while the rest of the sentence is normal. Research has shown that without being able to see the word at the center of fixation, people spend most of their time going back and forth between words, trying to guess what they are.
- The third is the Boundary paradigm. In this case, a selected target word in the sentence is changed to a preview before the current fixation reaches the word.
- Once the word is fixated, it is changed to its correct form. The preview could be a nonsense word, like the example shown here or could be a different word or anything else the experimenter can dream up.
- How the preview then affects the fixation time on the actual target word (compared to when the target word isn’t changed at all) indicates how much information is understood in the periphery and parafoveal as a person is reading.
- This figure shows an example sentence and where the invisible boundary is located.
- As a person’s fixation position crosses over the boundary, the program is triggered to change the target word.
- The preview could reveal nothing about the current word, the first few letters (while the rest of them are just random) or nothing at all about the target word (by substituting each letter with another letter, as in this case).