英语翻译4.The study4.1.Research questionsOur study examined three issues related to the successful adoption of SR for business applications.1.How accurate is text generated through SR as compared to typing?Prior studies have indicated SR accuraci

英语翻译4.The study4.1.Research questionsOur study examined three issues related to the successful adoption of SR for business applications.1.How accurate is text generated through SR as compared to typing?Prior studies have indicated SR accuraci
英语翻译
4.The study
4.1.Research questions
Our study examined three issues related to the successful adoption of SR for business applications.
1.How accurate is text generated through SR as compared to typing?
Prior studies have indicated SR accuracies up to
99%,but no tests have compared the accuracy of
subjects using SR and the keyboard to generate the
same text.Although most college-age students (our
subject population) know how to type,they vary in
their skill.Nevertheless,we expected the subjects
using SR to be less accurate than subjects using the
keyboard.H1.Subjects using SR are less accurate than subjects
using a keyboard.
2.How fast can text be generated through SR as
compared to typing?
A student typist may be able to type 13–41 wpm,
a good typist may type from 61 to 90 wpm,and an
excellent typist may type from 85 to 112 wpm,
assuming five characters per word[8].But people
*speak much faster than they type [32,34].When
limited by the necessity to generate free-form
comments in conversational English,people typi-
*cally speak at about 100–150 wpm [18].A second
study[20]found that a speaker with a normal speech
rate produces about 150 wpm,but the maximum
speech rate can be as much as 637 wpm[11].Thus,
we expected the subjects using SR to be able to
generate text faster than subjects using the keyboard.
*H2.Subjects using SR generate more text than sub-
jects using a keyboard.
3.What type of users will adopt speech recognition
technology?
Some people might not use computers because
of keyboard frustrations or an inability to under-
stand the interface.We expected users with less
keyboard experience would think that SR is easier
to use and more efficient.
*H3.Subjects with poor typing skills consider SR
easier to use and more efficient than subjects with
good typing skills.

英语翻译4.The study4.1.Research questionsOur study examined three issues related to the successful adoption of SR for business applications.1.How accurate is text generated through SR as compared to typing?Prior studies have indicated SR accuraci
4.研究
4.1 研究问题
我们的研究主要探讨了有关成功推广SR（计算机语音识别）商业应用的三个问题.
1 、与打字录入相比,声音识别录入的文本准确度如何?
以 前的研究表明,SR的准确度可高达百分之九十九,但从来没有进行就录入相同文档键盘录入跟语音录入所产生准确率的对比实验.虽然大多数的大学生（我们的研究对象）知道如何打字,但他们的能力各自不同.尽管如此,我们预期研究对象使用语音识别录入要比键盘录入准确率低.
*假想1：使用声音识别的调查对象要比使用键盘的调查对象录入准确率低.
2 、与打字录入相比,声音识别录入的速度如何?
一 个学生打字员也许每分钟能打13-41个单词 ,一个优秀打字员可能每分钟打61至90 个单词 ,一个出色的打字员可达到每分钟85-112个单词,在此假设每个单词的长度为五个字母 .而人们的说话速度远远超过打字的速度[ 32,34 ] .在对话英语中因为受到自由形态的交谈必要性限制,人们通常每分钟讲约100-150个单词[ 18 ] .另外一项研究[ 20 ]发现一名正常说话人的速度约是每分钟150个单词 ,但最高的讲话速度可达每分钟637个单词[ 11 ] .因此,我们预期采用声音识别录入的研究对象比使用键盘录入的研究对象能够以更快的速度输入文本.
*假想2：使用声音录入要比使用键盘录入输入文本多.
3 .哪类用户将采用语音识别技术?
使用键盘比较费劲或者因为键盘的界面不好理解,有些人可能就因而不使用电脑.我们预期有较少键盘使用经验的用户会认为声音录入更容易更有效.
*假想3：打字技巧较差的调查对象比打字能力强的调查对象认为声音录入使用起来更简捷.