Eadie Tanya L, Doyle Philip C
Department of Speech and Hearing Sciences, University of Washington, Seattle, 98105, USA.
J Voice. 2005 Sep;19(3):373-83. doi: 10.1016/j.jvoice.2004.04.004.
The purpose of this study was (1) to determine the psychophysical character of auditory-perceptual ratings of voice pleasantness (VP) and voice acceptability (VA) for tracheoesophageal (TE) speakers using direct magnitude estimation (DME) and equal-appearing interval (EAI) scaling procedures and (2) to determine the relationship between listeners' ratings of VP and VA. Ten adult listeners judged overall VP and VA from connected speech samples produced by 20 adult male TE speakers. Although results yielded a prothetic continuum for VP and a metathetic continuum for VA, the amount of variance accounted for by a curvilinear model of VP was minimally more than that accounted for by a linear model. Results also revealed a significant relationship between VP and VA (r = 0.939). Findings from this study do not suggest any greater validity associated with VP and VA ratings obtained by the DME than the EAI method. As a result of the significant relationship between these ratings and to the ease of applying EAI scales, it is recommended that VA be used as a current clinical outcome measure. These data illustrate the need to identify attributes that best describe TE speech that are measured appropriately and are clinically useful.
(1)使用直接数量估计(DME)和等距量表(EAI)程序,确定气管食管(TE)发音者语音愉悦度(VP)和语音可接受度(VA)的听觉感知评分的心理物理学特征;(2)确定听众对VP和VA的评分之间的关系。十名成年听众根据20名成年男性TE发音者的连贯语音样本判断总体VP和VA。尽管结果显示VP为单维连续体,VA为多维连续体,但VP曲线模型所解释的方差量仅略多于线性模型。结果还显示VP和VA之间存在显著关系(r = 0.939)。本研究结果并不表明通过DME获得的VP和VA评分比EAI方法具有更高的效度。由于这些评分之间存在显著关系,且EAI量表易于应用,因此建议将VA用作当前的临床结局指标。这些数据表明,需要确定最能描述TE语音的属性,这些属性应能得到适当测量且具有临床实用性。