This paper reports an experiment in which subjects were required to assess the grammatical accuracy of the spoken production of various speakers. Subjects were divided into groups and for each group one of the original voices was replaced by a computer-generated voice. A comparison was made between the mean scores given to the original voices and those given to the computer-generated ones. In this way the effect of different degrees of proficiency in pronunciation and fluency on the assessment of grammatical accuracy could be measured. It was found that there were noticeable differences in the mean marks given to the original and computer-generated voices, generally in favour of the original voices, whatever the level of pronunciation or fluency. Mean scores were also affected by the type of assessor: whether or not native speaker, and whether or not trained oral examiners.
Fluency and Pronunciation in the Assessment of Grammatical Accuracy in Spoken Production: An Empirical Study
CHRISTIANSEN, Thomas, Wulstan
2011-01-01
Abstract
This paper reports an experiment in which subjects were required to assess the grammatical accuracy of the spoken production of various speakers. Subjects were divided into groups and for each group one of the original voices was replaced by a computer-generated voice. A comparison was made between the mean scores given to the original voices and those given to the computer-generated ones. In this way the effect of different degrees of proficiency in pronunciation and fluency on the assessment of grammatical accuracy could be measured. It was found that there were noticeable differences in the mean marks given to the original and computer-generated voices, generally in favour of the original voices, whatever the level of pronunciation or fluency. Mean scores were also affected by the type of assessor: whether or not native speaker, and whether or not trained oral examiners.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
