Figure 2 compares the performance of patients and control subjects on iUFOV1 (identification task), iUFOV2 (dual task), and iUFOV3 (dual task with distractors). Note that higher scores represent worse performance in the intact hemifield (patients) or the left or right hemifield for controls. The scores discard performance in the contralateral field even though peripheral stimuli did appear in that hemifield during iUFOV2 and iUFOV3 (Methods). A mixed design ANOVA showed a significant interaction effect between group and subtest,
F(2,58) = 5.32,
PGG = 0.013. Contrasts showed a larger increase in presentation time from iUFOV1 to iUFOV2 in patients than for controls,
F(1,29) = 7.33,
P = 0.01. The difference between iUFOV2 and iUFOV3 was not different between patients and controls,
F(1,29) = 0.38,
P = 0.54. Post hoc tests showed that patients scored significantly worse than controls on iUFOV2 (mean
difference = 0.099 seconds,
P = 0.01) and iUFOV3 (mean
difference = 0.12 seconds,
P= 0.018), but not on iUFOV1 (mean
difference = 0.0013,
P = 0.17). Thus, patients needed longer presentation times to perform the double tasks than controls, but not the single identification task. In addition, patients’ scores were better for iUFOV1 than iUFOV2 (mean
difference = 0.112 seconds,
P < 0.001), which in turn were better than iUFOV3 (mean
difference = 0.081,
P = 0.003). For controls, however, we found no difference between iUFOV1 and iUFOV2 scores (mean
difference = 0.015 seconds,
P = 0.81). iUFOV3 scores were significantly higher than iUFOV2 (mean
difference = 0.062 seconds,
P = 0.01) and almost significantly higher than iUFOV1 (mean
difference = 0.077 seconds,
P = 0.05) for controls. Although we found no significant differences between the two groups’ mean and variance of age, we repeated the analysis with age as a covariate to ensure our results were not driven by age. This mixed design ANCOVA yielded similar results (
Supplemental Information S1). Age did not show any significant main or interaction effects.