LBNL Report Number
In a previous paper we analyzed Rea's data from a numerical verification (NV) experiment. We used a model based on discrete eye fixations, with fixation duration, composed of visual and nonvisual components. The present paper investigates one of the implications of this model, which involved an experiment on the effect of instructions on performance on the NV task. On any task requiring discrete eye fixations, the number of fixations needed for the task and the minimum time required per fixation puts a lower bound on the time to complete the task. The minimum time per fixation has been reported to be on the order of 200-300 ms for simple tasks. The two columns of numbers in the NV task are separated by about 7 degrees, which makes it difficult to clearly see them both in one fixation. On the other hand, a single five-digit number is only about 1.4 degrees across, so it should be visible in one fixation. This leads to an estimate of a minimum of 40 fixations to complete the task, and thus an estimate of 8-12 s as a minimum time per chart.