Effect Sizes Across Replications and Original (SNA96-1).
| Study . | Synth . | Venue . | Exclude Inattentives . | Nonword . | Partword . | Bupabi . | Non- v. Partword . | High- vs. LowTP . | Change- first v. -last . | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Original | MacinTalk | Uni | no | 76 | ** | 65 | ** | 74 | *** | 11 | * | 7 | * | 19 | ** |
| Exp. 1a | MBROLA | AMT | no | 61 | *** | 57 | *** | 52 | 4 | –1 | 1 | ||||
| Exp. 1a | MBROLA | AMT | yes | 65 | *** | 62 | * | 57 | 3 | 0 | 5 | ||||
| Exp. 1b | MBROLA | Uni | no | 67 | *** | 65 | *** | 68 | *** | 2 | 2 | 1 | |||
| Exp. 1b | MBROLA | Uni | yes | 70 | *** | 68 | *** | 71 | *** | 2 | 2 | 2 | |||
| Exp. 1c | WatsonTTS | AMT | no | 57 | 58 | NA | –1 | –1 | –8 | ||||||
| Study . | Synth . | Venue . | Exclude Inattentives . | Nonword . | Partword . | Bupabi . | Non- v. Partword . | High- vs. LowTP . | Change- first v. -last . | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Original | MacinTalk | Uni | no | 76 | ** | 65 | ** | 74 | *** | 11 | * | 7 | * | 19 | ** |
| Exp. 1a | MBROLA | AMT | no | 61 | *** | 57 | *** | 52 | 4 | –1 | 1 | ||||
| Exp. 1a | MBROLA | AMT | yes | 65 | *** | 62 | * | 57 | 3 | 0 | 5 | ||||
| Exp. 1b | MBROLA | Uni | no | 67 | *** | 65 | *** | 68 | *** | 2 | 2 | 1 | |||
| Exp. 1b | MBROLA | Uni | yes | 70 | *** | 68 | *** | 71 | *** | 2 | 2 | 2 | |||
| Exp. 1c | WatsonTTS | AMT | no | 57 | 58 | NA | –1 | –1 | –8 | ||||||
Note: Accuracies for nonword condition, partword condition, and the item bupabi, as well as differences in accuracies for the three main comparisons reported in SNA96-1. Green represents a successfully replicated effect across all replications. Standard errors, test statistics, and exact p-values are reported in the preprints.
* Significant at alpha level p = .05 under the revised analyses, [**] p = .01 under revised analyses, [***] p = .001 under revised analyses.