The other listeners were trained in a low-variability method—all of the sentences they heard came from a single speaker. Their task was simply to transcribe the sentences they were hearing. Following training, they completed a test phase that also involved transcribing sentences.