ENHANCING CHILDREN'S EVERYDAY COMMUNICATION: TALKER-SPECIFIC SPEECH RECOGNITION TRAINING

增强儿童的日常交流：特定于说话者的语音识别训练

基本信息

批准号：
9897507
负责人：
Nancy Ann Tye-Murray
金额：
$ 32.41万
依托单位：
WASHINGTON UNIVERSITY
依托单位国家：
美国
项目类别：
财政年份：
2016
资助国家：
美国
起止时间：
2016-04-01 至 2022-03-31
项目状态：
已结题

来源：
https://reporter.nih.gov/project-details/9897507
关键词：
Academic achievement Accountability Address Adult Advocate Affect Age Auditory Belief Characteristics Child Clinic Clinical Cognitive Science Communication Computer software Computers Couples Data Depressed mood Devices Education Environment Exercise Experimental Designs Exposure to Future Teacher Goals Hand Hand functions Hearing Aids Home environment Hour Human Individual Individual Differences Internet Laboratories Language Language Development Learning Light Link Methods Modeling Motivation Outcome Outcome Measure Participant Pattern Performance Protocols documentation Research Sampling Schools Self-Direction Signal Transduction Socialization Specialist Speech Speech Perception Spouses Structure Students System Testing Time Training Training Programs Universities Visual Voice Work audiovisual speech aural rehabilitation base comparative efficacy compliance behavior computerized cost effective deaf design evidence base hearing impairment high risk improved improved outcome innovation multidisciplinary normal hearing programs public health relevance social speech in noise speech recognition teacher theories

项目摘要

DESCRIPTION (provided by applicant): Almost 90% of children with hearing loss attend a regular public school classroom and 95% communicate with spoken language. Depending upon their degree of hearing loss, they miss 30-60% of all spoken material in a classroom setting, even if they wear hearing aids. Not surprisingly, research has shown that poor speech recognition leads to poor academic achievement; it is a formidable task to learn academic material if one understands only three in every ten words. Although most experts in aural rehabilitation advocate structured training to improve speech recognition, the sad reality is that most children do not receive it. This project offers a potential solution to this conundrum. Using a feasible and inexpensive delivery system, our multi-disciplinary team proposes to determine the kinds of speech recognition training that most benefit children. Hand-held, game-like training programs will be used to present speech-in- noise training to three groups of children who have hearing loss: one group will receive auditory training, one will receive audiovisual training, and one will receive a combination of both. Following an A-A-B-B-b experimental design, group benefits will be compared after 4 weeks of formal training in a clinic-like setting, after 4 weeks f required home training, and then after semester-long self-determined training. Importantly, using a Transfer-Appropriate Processing (TAP) theory of learning as theoretical motivation, we will determine the extent to which children benefit from talker-specific training. Upon completion of the project, our team will have determined which type of speech recognition training results in the greatest gains in speech recognition (Aim 1), whether we can tailor speech recognition training to quickly integrate children into new classrooms by having them train with materials spoken by a future teacher (Aim 2),and the extent to which home-based training via hand-held devices can maintain and expand upon gains made in a clinic-like training environment and how patterns of training and individual participant variables affect benefit (Aim 3). The project incorporates innovation on multiple levels, including the comparison of training methods, the theoretical substrate, the test battery and consideration of visual enhancement, the incorporation of game design principles into program software, and the experimental design. Improving speech recognition using evidence-based training offers the best and most practical solution to improving academic performance for children who are at high risk for scholastic and social difficulties. The proposed research, which is based firmly on TAP theoretical underpinnings, will provide answers about how best to provide speech recognition training to children with hearing loss.

描述（由申请人提供）：几乎 90% 的听力损失儿童在正规的公立学校课堂上上课，95% 的儿童使用口语进行交流，根据听力损失的程度，他们错过了课堂上 30-60% 的口语材料。不足为奇的是，研究表明，语音识别能力差会导致学业成绩不佳；如果一个人只能理解十分之三的单词，那么学习学术材料就是一项艰巨的任务。提倡结构化培训来提高语音识别能力，但可悲的现实是大多数孩子没有接受它，该项目为这个难题提供了一个潜在的解决方案，我们的多学科团队建议使用可行且廉价的传输系统来确定语音类型。最有利于儿童的识别训练将使用类似游戏的手持式训练程序向三组听力损失儿童进行噪声中言语训练：一组将接受听觉训练，一组将接受视听训练，一个人将收到两者的组合。 A-A-B-B-b 实验设计，在类似诊所的环境中进行 4 周的正式培训后，在需要的 4 周家庭培训后，然后在为期一学期的自主培训后，将比较小组的效益，重要的是，使用适当的转移处理（ TAP）学习理论作为理论动机，我们将确定儿童从特定说话者训练中受益的程度。项目完成后，我们的团队将确定哪种类型的语音识别训练可以带来最大的语音收益。识别（目标 1），我们是否可以定制语音识别训练，让孩子们使用未来老师所说的材料进行训练，从而快速融入新课堂（目标 2），以及通过手持设备进行家庭培训的程度可以维持和扩展在类似诊所的培训环境中取得的成果，以及培训模式和个人参与者变量如何影响效益（目标 3）。该项目纳入了多个层面的创新，包括培训方法的比较、理论基础、测试电池并考虑视觉增强，将游戏设计原理纳入程序软件以及使用基于证据的训练改进语音识别为提高学业和社交困难风险高的儿童的学业成绩提供了最佳和最实用的解决方案。该项目牢固地基于 TAP 理论基础，将提供有关如何最好地为听力损失儿童提供语音识别训练的答案。