Figure 4

Signal reliability for Experiment 2 expressed as t-statistics (Mean ± SD). The regressive method for determining the response to the stimuli meant that each of the 176 sensitivity measurements/subject (2 eyes × 2 pupils × 44 regions/eye) has its own SE. Thus, a univariate t-statistic testing the case of no response could be computed, to quantify the reliability of the 176 pupil responses. The plotted values are the means around the 5 rings of the stimulus arrays for the 23 subjects of Experiment 2. The abscissa gives the eccentricity of the centre of each ring. The symbols for the No-Attention and Attention cases are slightly offset to the left and right to allow the symbols and error bars to be seen. The signal reliability was uniformly high in the peripheral regions where the effects of attention were most evident.