COMPUTER-SPEECH FEEDBACK IN TEXT FOR DYSLEXIC CHILDREN

针对诵读困难儿童的计算机语音文本反馈

• 批准号: 2025150
• 负责人: RICHARD K OLSON
• 金额: $ 29.2万
• 依托单位: UNIVERSITY OF COLORADO
• 依托单位国家: 美国
• 财政年份: 1987
• 资助国家: 美国
• 起止时间: 1987-01-01 至 1998-12-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/2025150
• 关键词: auditory discrimination; auditory feedback; behavioral /social science research tag; child psychology; clinical research; computer assisted instruction; computer graphics /printing; computer human interaction; computer system design /evaluation; cues; human subject; language development; longitudinal human study; middle childhood (6-11); reading; reading disorder; remedial /special education; speech; speech synthesizers; training aid

The proposed research builds on previous and current studies of computer- based remediation for reading-disabled children, with a new comparison of two theoretically and practically different ways of training phoneme awareness. The story reading portion of the training will be similar to previous grants. In each training year of the proposed study, 108 second- to fifth-grade students with reading disabilities will read interesting stories on a computer for 30 minutes daily during their remedial-reading or language-arts class time. They will target difficult words in the stories with a mouse for immediate orthographic and speech feedback. The computer will emphasize important relations between sub-word letter patterns and speech sounds, by segmenting the word into helpful units for the student to sound out, and then by pronouncing each segment in order while highlighting the segment. Previous studies revealed significant gains in phonological decoding, word recognition, and attitude about reading for reading-disabled students trained on the computer compared to similar students receiving normal classroom or remedial reading instruction. However, trained students with relatively lower phoneme awareness had much smaller average gains than those with higher phoneme awareness. Current studies show that supplementary, articulatory-based training in phoneme and phonological awareness on the computer can reduce or erase these deficits and support substantially greater gains in word decoding. Now we propose to compare two different phonological-awareness programs and an untrained control group. Both training conditions will include intensive practice in manipulating phonemes and graphemes in reading and spelling exercises, but one will pretrain articulatory awareness while the other pretrains sound-letter based phoneme awareness, to isolate unique effects of concrete speech-motor awareness. We hypothesize that training in articulatory awareness will be most critical for children with the weakest initial phoneme awareness, while other children with less severe reading disabilities may benefit as much from non-articulatory based training in phoneme awareness and phonological decoding. We will also explore the addition of visual speech cues from an animated mouth and face that coordinates with the synthetic speech. Longitudinal assessment through individual growth-curve analyses will describe subjects' development of reading and language skills during and up to 1 year after training.

拟议的研究建立在以前和目前对计算机的研究基础上- 以阅读障碍儿童为基础的矫正，与新的比较 两种理论和实践上不同的音素训练方法 意识。培训的故事阅读部分将类似于 以前的拨款。在拟议研究的每个培训年度，108秒-- 对于有阅读障碍的五年级学生来说，阅读会很有趣 每天在电脑上阅读30分钟的故事 或者是语言艺术课时间。他们将针对较难的单词在 使用鼠标即时获得拼写和语音反馈的故事。这个 计算机将强调字母子词之间的重要关系 模式和语音，通过将单词分割成对 让学生发音，然后按顺序读出每个片段 同时突出显示细分市场。此前的研究表明， 在语音解码、单词识别和对 在计算机上训练的阅读障碍学生的阅读能力与 接受正常课堂或补习阅读的相似学生 指示。然而，训练出音素相对较低的学生 与那些音素较高的人相比，认知能力的平均增长要小得多 意识。目前的研究表明，辅助性的、基于发音的 在计算机上进行音素和语音意识训练可以减少 或者消除这些不足，并在Word中支持更大的收益 解码。现在我们来比较两种不同的语音意识 程序和未受过训练的控制组。这两个培训条件都将 包括在操作音素和字素方面的密集练习 阅读和拼写练习，但要预先训练发音能力 意识，而另一个预先训练基于音字母的音素意识， 以分离具体言语运动意识的独特效果。我们 假设发音意识的训练将是最关键的 对于初始音素意识最弱的儿童，而其他 阅读障碍不那么严重的儿童可能会从 基于非发音的音素意识和音位训练 解码。我们还将探索添加来自 与合成语音协调的动画嘴巴和脸部。 通过个人成长曲线分析进行纵向评估将 描述受试者阅读和语言技能的发展。 培训后最多一年。