Hacker News Reader

Comment by kristiandupont | Hacker News Reader

Does anybody know of studies correlating eye movement with mouse position? All these tools sort of pretend to be revealing eye movement but I find that I personally keep the mouse cursor in one position, scroll and only hover a link if I have already decided to click it..

It's obviously still valuable data, but just not quite the mind reading super tool that it pretends to be.

View on HN

This paper does, and also tries to predict where you are looking based on cursor activity, rather than just rely on the cursor's current location:

http://jeffhuang.com/Final_GazeCursor_CHI12.pdf

"But claiming that the cursor approximates the gaze is misguided - as we have shown, this is often not the case depending on time and behavior"

"Therefore, we conclude that whether the subject tracks their gaze closely with their cursor is more likely to stem from personal habits rather than age or gender."

I did a bit of research a few years ago as part of my masters. Here are the references I have:

http://stc-access.org/wp-content/uploads/cooke_mouse_eye_tra...

TLDR: 'eye movement matched mouse movement 69% of the time'

There's also a paper in here starting page 33 of the PDF http://research.microsoft.com/en-us/um/people/ryenw/proceedi...

Yeh the data is valuable if used in a meaningful way. You'd still want to A/B test different things you change but our tool would give you a good starting point based on user/visitor behaviour.

Reply on news.ycombinator.com