Visual search display

(x=0,y=0)

Trial 3 task: find

y

(Top-right)

(x=1024,y= 0)

0

Screen center:

(512, 384)

384

(1024, 768)

(Bottom right)

768

0

512

1024

x-coordinate (pixels)

Any patterns in search? To interpret the time series (below) we need to know about the screen coordinates (above). These might seem a bit counterintuitive. Note (above) the x range is 0 to 1024 from lft to rt, and the y range is 0 to 768 top to bottom. X & Y screen dimensions have (0,0) coordinates at the top left corner of the screen, (512, 384) is the center, and (768, 1024) at the bottom right. Thus, When the tracked eyes move to the left (#s decrease on x-axis) they tend to also move lower in the screen (#s increase on y-axis). Additional search patterns are described below.

X dec= left mvmt

y dec= upward mvmt**

1) Left moving eyes tend to also move down. This is easily seen above in the (apparently) opposing light-blue & red patterns. These illustrate the correlation between x & y directions with left-downward mvmt tending to co-occur as do right-upward movements. Perhaps this has to do with left to right reading bias; but this is pure speculation.

2) Velocity units are visual degrees/second & shown in pink &purple at bottom of figure above. Some x & y contrast again appears with increases in velocity in the horizontal direction tend to be associated with decreases in the vertical direction. Another trend is the larger range, variability and magnitude of velocity in horizontal eye-movements. This can be attributed to the asymmetry in the shape of our eyes permitting greater movement along the horizontal dimension. Of greater significance is *the change in behavior* as a function of change in the size of these spatial dimensions. This pertains to the important fractal characteristic of scaling present in many natural systems.

3) What about all that variability? The main point of the present study is its focus on the variability present in eye-movements. Here a key question is whether this is random variability or is there some structure? When we ask about whether there is structure we are interested in patterns across the time series, or whether a statistical memory persists over the time series. As we'll see, we will be asking quesitons about what kind of noise is present in the data and whether or not it is considered a brownian or colored form of noise & finally can we detect fractal patterns? Here is some more information on these ideas.