Table 2.
Summary of GEE Model Effects including long-term user group (high and low users) and visit (12 months, 18 months, and 3.5 years), and task half (first, second) as predictors of the Anti-saccade Task outcome measures.
| Wald χ2 (df), p-value | |
|---|---|
| % Anti-saccades | |
| Half | 125.02 (1), p < 0.001 |
| Visit | 10.80 (2), p = 0.005 |
| Group | 0.14 (1), p = 0.706 |
| Half * group | 1.30 (1), p = 0.254 |
| Visit * group | 2.00 (2), p = 0.369 |
| Half * visit | 5.46 (2), p = 0.065 |
| Half * visit * group | 0.82 (2), p = 0.661 |
| % Pro-saccades | |
| Half | 230.34 (1), p < 0.001 |
| Visit | 10.03 (2), p = 0.007 |
| Group | < 0.01 (1), p = 0.966 |
| Half * group | 0.03 (1), p = 0.871 |
| Visit * group | 1.49 (2), p = 0.476 |
| Half * visit | 10.56 (2), p = 0.005 |
| Half * visit * group | 0.13 (2), p = 0.938 |
| % Corrective looks | |
| Half | 5.01 (1), p = 0.025 |
| Visit | 14.29 (2), p = 0.001 |
| Group | 3.26 (1), p = 0.071 |
| Half * group | 9.35 (1), p = 0.002 |
| Visit * group | 1.37 (2), p = 0.505 |
| Half * visit | 6.52 (2), p = 0.038 |
| Half * visit * group | 1.08 (2), p = 0.583 |
| Latency to distractor (pro-saccade) | |
| Half | < 0.01 (1), p = 0.993 |
| Visit | 17.70 (2), p < 0.001 |
| Group | 4.55 (1), p = 0.033 |
| Half * group | 1.75 (1), p = 0.186 |
| Visit * group | 4.96 (2), p = 0.084 |
| Half * visit | 9.06 (2), p = 0.011 |
| Half * visit * group | 1.69 (2), p = 0.430 |
| Latency to target location (anti-saccade) | |
| Half | 9.28 (1), p = 0.002 |
| Visit | 33.69 (2), p < 0.001 |
| Group | 0.94 (1), p = 0.334 |
| Half * group | 1.71 (1), p = 0.191 |
| Visit * group | 0.26 (2), p = 0.878 |
| Half * visit | 4.04 (2), p = 0.133 |
| Half * visit * group | 0.79 (2), p = 0.673 |
The analysis included 14 LUs and 24 HUs.
Significant results are in bold.