Equating character-identification performance across the visual field.

The effect of retinal eccentricity on stimulus identification was investigated by measuring observers' ability to identify digits of various sizes presented briefly at different locations in their visual fields. The stimuli were either presented one at a time or presented in pairs with each member of the pair at a different eccentricity and having a different size. Stimulus-response confusion matrices were used to calculate the information transmitted by each character as a function of stimulus size and retinal eccentricity. Linearly scaling stimulus size with retinal eccentricity yields equal information transmission across each subject's visual field. There was no difference between the amounts of information transmitted when the targets were presented alone and when they were transmitted in pairs. For small target sets, therefore, the amount of information transmitted from the screen to the observer can be approximated as the sum of the information at the separate locations.