17.May.2010 A little knowledge…

… is a dangerous thing, and today proves this. When you start reading about eye movement (and this is my core PhD topic), a lot of texts go like: ok, so the eye movement is made up of jumps (called saccades) and times when the eye is stationary (fixations). You also have backward jumps (regressions) and things like return sweeps (end of line to the beginning of the next). When you actually do eye-tracking, or read the more detailed articles, it turns out that the eye is not really stationary during a fixation but is actually always moving. So you discover that there’s things called tremors, microsccades, and drifts… OK, not a problem.

The thing that many articles don’t tell you is that the eyes are not really fixating on the same spot except half the time (aligned around 55% of the time). The other times they are more than 1 character apart in fixation position, and sometimes they are even crossed (left eye fixation on the right of the right eye fixation). Crazy? Oh, well… It seems this is why most studies are recording only 1 pupil and not both. Why is this bugging me? My eye tracker recorded both pupils and I have already extracted all that data. I might have to extract all that again, or try to see of the average of the 2 pupil positions is a viable thing to study. Thankfully, I had not started the data analysis, or I would be crying now. Disaster averted, till now at least. Yesterday I wrote on Facebook about my PhD: It’s like I have to climb Mount Everest and I have no shoes on! Hmm… Shopping really is the answer to everything :)