Using E-Z Reader to examine the consequences of fixation-location measurement error.

Area of Science:

• Cognitive psychology
• Computational neuroscience
• Psycholinguistics

Background:

• The debate on reading eye movements centers on whether attention is strictly serial or gradients.
• Parafoveal-on-foveal and spillover effects are often cited as evidence for attention gradients.

Purpose of the Study:

• To investigate if measurement errors in fixation locations can simulate attention-gradient effects.
• To test the robustness of the serial-attention hypothesis against observed reading phenomena.

Main Methods:

• Three simulations were conducted using the E-Z Reader computational model of eye-movement control.
• Systematic and random errors in fixation location measurements were introduced.
• The model's output was analyzed for parafoveal-on-foveal and spillover effects.

Main Results:

• Simulations demonstrated that measurement errors can artificially generate parafoveal-on-foveal effects.
• Large spillover effects were also observed as a consequence of simulated measurement inaccuracies.
• These effects occurred even when the model strictly adhered to serial attention allocation.

Conclusions:

• Parafoveal-on-foveal and spillover effects do not necessarily refute the serial-attention hypothesis.
• Measurement error is a critical factor to consider when interpreting eye-movement data in reading research.
• The findings suggest a re-evaluation of evidence previously used to support attention-gradient models.