For saccade amplitudes, a repeated-measures ANOVA found a main effect of visual condition (
F(1.38,56.7) = 237.002,
p < 0.001, η
2p = 0.85) and a main effect of time interval (
F(1,41) = 31.46,
p < 0.001, η
2p = 0.43), qualified by a significant interaction (
F(2,82) = 5.31,
p = 0.007, η
2p = 0.12). Simple effects tests (α = 0.017) indicated significant effects of time interval for the peripheral vision (
F(1,41) = 21.4,
p < 0.001, η
2p = 0.34) and the control conditions (
F(1,41) = 24.1,
p < 0.001, η
2p = 0.37), whereas the effect for the central vision condition was not significant (
F(1,41) = 5.82,
p = 0.020, η
2p = 0.13). As predicted, saccades during early time intervals were significantly shorter than saccades during late time intervals for the control and peripheral visual conditions, whereas the effect for the central vision condition was in the same direction but non-significant post-Bonferroni correction.
Figure 6B depicts mean saccade amplitudes in early and late time intervals for each of the three visual conditions. Planned contrasts indicated that, as predicted, during the early time interval, the peripheral vision condition (
M = 9.3 degrees) involved larger saccade amplitudes than the control condition (
M = 6.4 degrees), which in turn were larger than the central vision condition (
M = 3.9 degrees), all
p values < 0.001. An identical pattern occurred for the late time interval, with the peripheral vision condition (
M = 8.5 degrees) involving larger amplitudes than the control condition (
M = 5.8 degrees), which in turn were larger than the central vision condition (
M = 3.7 degrees), all
p values < 0.001. These results were consistent with the hypothesis that the peripheral vision condition should involve larger saccade amplitudes than the control condition, which in turn should involve larger saccades than the central vision condition, due to differences in the information available to the visual system for saccade planning in the three conditions. Furthermore, it is consistent with the hypothesis that such differences in saccade amplitudes between the three conditions should be accentuated during the early time interval, when the ambient mode of vision is active.