Abstract
Purpose :
In the pro- and anti-saccade task participants are required to shift their gaze towards an appearing target (pro-saccade) or away from it to the opposite direction (anti-saccade). Pro-saccades are considered reflexive and hard to inhibit while anti-saccades are volitional. Despite the widespread use and utility of these tasks they still lack accepted standardized norms, partially because each lab adopts its own specific protocol. A fundamental consideration is whether to separate the anti- and pro- conditions into blocks, or to randomly interleave them within one larger experiment. We hypothesized that administration will have an effect on anti-saccade performance. Results surprisingly revealed that pro-saccades were also affected.
Methods :
21 right-handed undergraduate students completed block and interleaved versions of the pro and anti saccade task. Each trial began with a 2000ms fixation at center screen followed by a 600ms oval target that appeared (randomly and with equal probability) left or right of fixation. In interleaved trials fixation also served to indicate the saccade type: a plus sign for pro and a circle for anti. Monocular eye movements were recorded at 120 Hz using the ISCAN ETL-400 video system.
Results :
Repeated measure ANOVA's confirmed that overall pro-saccades have shorter latency (M=254ms) and smaller error rates (M=8%) compared to anti-saccades (M=326ms; M=32%). More interestingly, interleaved mode resulted in larger error rates not only for anti (blocked:M=22%; interleaved: m=42%) but also for pro saccades (blocked: M=5%; interleaved: M=12%). This difference between block and interleaved administration was significantly larger in anti compared to pro-saccades. *all smaller than p<.01.
Conclusions :
As hypothesized interleaved trials resulted in larger error rates for anti-saccades. Surprisingly, it also increased pro-saccade error rates. If pro-saccades were indeed solely reflexive, then they should not be affected by test administration. Our results indicate that this is not the case; rather administration procedure does affect pro-saccades. Instead of a binary classification of reflexive pro and volitional anti-, we suggest a continuous model where both types of saccades can be placed. Anti-saccades require more inhibition and are hence more, but not entirely, volitional while pro-saccades are less taxing on cognitive resources and hence are more, but not entirely, reflexive.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.