Contrast efficiencies for different ER sequences of five fingertip stimulations
| Contrast | Fast ER (selected) | Fast ER (average) | Slow ER (average) | Fast ER (fully randomized) | Fast ER (no null event) |
|---|---|---|---|---|---|
| Detection (fingertip vs baseline) | 4.29 ± 0.49 | 2.95 ± 0.25 | 1.38 ± 0.09 | 2.41 ± 0.17 | 1.45 ± 0.03 |
| HRF estimation | 0.43 ± 0.02 | 0.43 ± 0.03 | 0.20 ± 0.02 | 0.38 ± 0.03 | 0.03 ± 0.00 |
| Fingertip difference detection | 2.77 ± 0.25 | 2.49 ± 0.22 | 0.97 ± 0.03 | 1.70 ± 0.10 | 1.82 ± 0.12 |
The leftmost column gives the average efficiencies for the fast ER sequences actually used in this study, selected among the 20 sequences with the largest detection efficiency (out of 50,000 randomly drawn sequences). The second and third columns give the average efficiency of randomly drawn sequences for the fast and slow ER designs. The fourth and fifth columns give the efficiencies of modified fast ER sequences, one in which events were randomized across the entire sequence instead of blocks of 21 events (fourth column) and another that did not include null events (fifth column). The three rows, respectively, correspond to overall efficiencies for detection contrasts (fingertip stimulation vs baseline, assuming a canonical HRF function), HRF estimation and difference detection contrasts (pairwise fingertip stimulation differences, also assuming a canonical HRF function). All efficiencies are given with their respective standard deviations across 20 or 50,000 sequences.