These materials were designed for familiarising participants with three new talkers then subsequently testing the intelligibility of sentences spoken by trained and novel voices in the presence of a competing sentence.
Speech stimuli are sentences spoken by five talkers. Open-set (high predictability) sentences are used for familiarisation and training, and closed-set (BUG matrix) sentences are used for intelligibility testing. For each participant, three voices are (pseudo-randomly) selected for training, and the other two voices are selected as novel voices only presented during the speech intelligibility test.
During familiarisation, participants hear sentences spoken by three talkers and a name appears on the screen, which participants are instructed to associate with the voice.
During training, participants complete a voice identification test with the three voices presented during familiarisation: They hear a sentence, then are asked which of the three names the voice corresponds to. They respond by clicking one of three buttons on the screen, corresponding to the three possible name words. Participants receive feedback about whether their response was correct or incorrect.
During the speech intelligibility test, participants hear a sentence in the presence of a competing sentence. They are instructed to listen to the sentence that begins with a particular word and report the remaining four words from the target sentence, by clicking one word from each column on the screen (in any order). On some trials, the target sentence is spoken by one of the three trained (i.e., familiar) voices and the masker sentence is spoken by one of two novel voices. On other trials, both sentences are spoken by (different) novel voices.
Built with Experiment
These materials were designed for familiarising participants with three new talkers then subsequently testing the intelligibility of sentences spoken by trained and novel voices in the presence of a competing sentence.
Speech stimuli are sentences spoken by five talkers. Open-set (high predictability) sentences are used for familiarisation and training, and closed-set (BUG matrix) sentences are used for intelligibility testing. For each participant, three voices are (pseudo-randomly) selected for training, and the other two voices are selected as novel voices only presented during the speech intelligibility test.
During familiarisation, participants hear sentences spoken by three talkers and a name appears on the screen, which participants are instructed to associate with the voice.
During training, participants complete a voice identification test with the three voices presented during familiarisation: They hear a sentence, then are asked which of the three names the voice corresponds to. They respond by clicking one of three buttons on the screen, corresponding to the three possible name words. Participants receive feedback about whether their response was correct or incorrect.
During the speech intelligibility test, participants hear a sentence in the presence of a competing sentence. They are instructed to listen to the sentence that begins with a particular word and report the remaining four words from the target sentence, by clicking one word from each column on the screen (in any order). On some trials, the target sentence is spoken by one of the three trained (i.e., familiar) voices and the masker sentence is spoken by one of two novel voices. On other trials, both sentences are spoken by (different) novel voices.
There are 10 voice counterbalances, such that different participants are trained with a different subset of 3 voices from the set of 5 voices. The remaining 2 voices from the set are presented as unfamiliar.
NOTE: The original shared version of this experiment contained an error, such that the counterbalance condition for each participant did not affect the stimuli they were presented with in the speech intelligibility test, and the condition names in the speech intelligibility data file were incorrect. This has been corrected in the current version: The counterbalance condition for each participant now correctly determines the stimuli that participants are exposed to during familiarisation and training, and the stimuli they are presented with during the speech intelligibility test
Fully open! Access by URL and searchable from the Open Materials search page
