Wilson R H, Arcos J T, Jones H C
J Speech Hear Res. 1984 Sep;27(3):378-86. doi: 10.1044/jshr.2703.378.
Consonant-vowel-consonant (CVC) monosyllabic words were segmented at the approximate phoneme boundaries and were presented to subjects with normal hearing in the following sequence: (a) the carrier phrase to both ears, (b) the initial consonant segment to one ear, (c) the vowel segment to the other ear, and (d) the final consonant segment to the ear that received the initial consonant. A computer technique, which is described in detail, was used to develop the test materials. The digital editing did not alter appreciably the spectral or temporal characteristics of the words. A series of four experiments produced a list of 50 words on which 10% correct word recognition was achieved by listeners with normal hearing when the vowel segment or the consonant segments of the words were presented monaurally in isolation. When the speech materials were presented binaurally--that is, the vowel segment in one ear and consonant segments in the other ear--word-recognition performance improved to 90% correct.
辅音-元音-辅音（CVC）单音节词在大致的音素边界处进行分割，并按照以下顺序呈现给听力正常的受试者：（a）向双耳呈现载体短语；（b）向一只耳朵呈现初始辅音片段；（c）向另一只耳朵呈现元音片段；（d）向接收初始辅音的耳朵呈现末尾辅音片段。使用一种详细描述的计算机技术来开发测试材料。数字编辑并未明显改变单词的频谱或时间特征。一系列四个实验产生了一个包含50个单词的列表，当单词的元音片段或辅音片段单独单耳呈现时，听力正常的听众对这些单词的正确识别率达到10%。当语音材料双耳呈现时，即一只耳朵呈现元音片段，另一只耳朵呈现辅音片段，单词识别性能提高到90%的正确率。
