Methodological variables in choral reading [corrected] [published erratum appears in CLIN LINGUISTICS PHONETICS 2008 Feb;22(2):169].

This preliminary study explored changes in prosodic variability during choral reading and investigated whether these changes are affected by the method of eliciting choral reading. Ten typical adult talkers recorded three reading materials (poetry, fiction and textbook) in three reading conditions: solo (reading aloud alone), track (reading aloud with a recording), and choral (reading aloud with another live talker). Measurements of fundamental frequency, amplitude, and vowel duration variability were performed. Compared to solo reading, choral reading featured decreased variability of fundamental frequency, amplitude, and vowel duration. In track reading, only decreased fundamental frequency variability was observed, while vowel duration variability increased. Track reading also contained significantly more errors than choral or solo reading. The reading material used did not significantly affect the prosodic features of talkers' speech, but did affect talkers' error rate. These results suggest that, in studies of choral reading as a fluency-evoking condition, track reading might not be an appropriate substitute. In addition, the text used may affect fluency in these conditions. These results also have theoretical implications for studies of how talkers manipulate and execute the prosodic features of speech. [ABSTRACT FROM AUTHOR]