An Experimental Study On The Effects Of Voice-to-Text Software On MTI Students' C-E Consecutive Interpreting Performance

Voice-to-text software is likely to have the ability to improve interpreters' performance.This report presents an experimental study that tests the accuracy of voice-to-text software and its effects on MTI students' C-E consecutive interpreting performance.According to previous studies,voice-to-text software becomes more and more popular in many situations,such as during classes,exhibitions,speeches,etc,due to its high accuracy.However,voice-to-text software is not widely used in interpreting.The author makes a consumption that it is also useful in consecutive interpreting.The author does this experiment to see its effects on students' C-E consecutive interpreting performance.In the first stage of the experiment,the author tests the accuracy of Speechlogger,a kind of voice-to-text software,in transcribing two Chinese recordings.It is accepted that voice-to-text software performs poorly in transcribing English.Therefore,the author does not bother to test its accuracy in transcribing English.In the second stage,the author tests the effects of voice-to-text software on MTI students' C-E consecutive interpreting performance.20 MTI students,whose interpreting skills are of similar level,are divided into two groups.They are required to interpret two Chinese recordings into English.These two recordings are different in topics and number of numbers,coordinate structures and terms.Group 1 interpret the recordings without the help of voice-to-text software;while Group 2 interpret the recordings with the aid of voice-to-text software.The author scores their performance from the perspectives of fluency,completeness and accuracy of coordinate structures,terms,and numbers.After the experiment,the author asks students to give some feedback on the experiment.According to the result of the experiment,the author finds voice-to-text software is of high accuracy rate in transcribing Chinese,over 95%.Furthermore,students who interpret with the help of transcriptions score higher than those who do not on the whole in both recordings.Voice-to-text software can improve the completeness and accuracy of coordinate structures,terms,and numbers,especially for materials that have plenty of complex coordinate structures,terms,and numbers.However,their fluency will be affected for they sometimes refer to the transcriptions.