Idris,

Fascinating question. It is plausible to say that saccades indicate confusion but you would need a normative definition of "non-confusion" before you can define confusion.

Zach

Hi,

keep in mind that eye movements may reflect several different cognitive processes, that potentially range from object identification over information retrieval (from memory) to monitoring. One fixation registered on an object at t1 could reflect process X and the next fixation on the same object registered at t2 could reflect process Y. Unless you know what process you're measuring at what point in time, it is difficult to argue a regression reflects confusion or difficulty.

Johannes

@zachary & @Johannes. Thanks for your response. 

That makes sense.

What I am thinking is in a visual search where a subject is tasked with identifying targets among distractors (both resembling one another to varying degrees), what would multiple re-fixations on an object signify? Say an object 1 receives significantly large re-fixations compared to object 2 before it was identified, then how plausible would it be to argue that object 1 may have characters or features that are comparatively difficult to process

Still, very difficult to directly answer that. There are too many unknown factors for me. But I found the work by Ian Gilchrist very interesting and enlightening in my work on the role of attention during visual processing (not only search). Could be that this article gives you the answers you're looking for: http://www.sciencedirect.com/science/article/pii/S0960982200007296 

Johannes

@ Idris

Idris,

In second language reading, returning to an object area (1-2 words) can indicate mis-comprehension, or simply that the eye skimmed passed during the saccade. In a child eye tracking experiment I was part of, large re-fixations indicated comprehension.

Johannes is correct that there are too many unknown factors. I think if you are able to operationally define confusion and non-confusion, you will have a baseline with which to conduct and experiment and start to formulate ideas about what phenomena may be occurring. 

Very interesting stuff. 

Zach

Dear Idris,

I agree with the preceding comments.   Your research idea is very interesting, but you would need to separate causation.   For example, you would have to isolate perceptual factors (e.g. processes from the retina to the visual cortex including the mid-brain/ metencephalon) from cognitive factors (e.g. from Wernicke's area and beyond).   Additionally, your experimental material would need to be able to separate bottom-up from top-down drivers.

That being said, I think that this research would be very interesting and could divulge some highly useful data/theories.   If you are interested, I can send you some material that will assist you to lay a foundation for investigating this question.

Wishing you all the very best,

Bruce Hilliard

@ Johanson :Thanks for your response and the link. The article is was quite informative.  I agree, there are too many unknown parameters to draw any conclusions at this point.  

@Zach. Thanks for your reply. I am thinking to measure cognitive load (using pupil dilation or EEG) and use that as a baseline  and try to find  a corelation between cognitive load and regressive saccades. My initial test runs indicate there might be but I am not very sure. (I am an engineer and have a  limited background in this, so I could be catastrophically wrong!)

@Bruce: Thanks for your comments. Like I said I am just beginning to explore the area, so any reading material would be of great help. Thanks

Idris ~ Give yourself credit. You are asking very good questions and your baseline measures should give you a solid foundation. Thanks for the conversation. 

Dear Idris,   Contact me through [email protected] and I will send you the background reading material that can help you to lay a foundation for your research.

Best Regards

Bruce Hilliard