Table 5.
Information transfer estimates for consonant features in 24 cochlear implant users from van Wieringen and Wouters (1999). Next to each feature in parentheses is the number of feature categories. Avg: average estimate per user, N=192 samples per matrix; Pool8: estimate obtained by averaging the pooled estimate between better, intermediate, and poorer performers (8 listeners per group with N=1536 samples per matrix); and Pool24: estimate obtained by pooling data from all 24 listeners into one matrix with N=4608 samples. Without the small-sample bias, numbers in a given row should be similar.
| (%) | Consonants | ||
|---|---|---|---|
| Avg | Pool8 | Pool24 | |
| Voicing (2) | 41 | 35 | 33 |
| Burst (2) | 57 | 53 | 51 |
| Amp Env (2) | 63 | 59 | 56 |
| Nasal (2) | 43 | 41 | 37 |
| Affrication (2) | 52 | 46 | 43 |
| Manner (3) | 58 | 53 | 49 |
| Place (4) | 15 | 10 | 8 |