Schissel R J, Flournoy J E
J Commun Disord. 1978 Sep;11(5):459-68. doi: 10.1016/0021-9924(78)90039-4.
This study examines the extent to which the mean, or average degree of agreement, of a group of listeners actually represents the consistency of judgments of all members of that group. Twenty-two experienced and inexperienced judges scored the responses, from tape recordings, of six children with multiple articulation errors on the McDonald Screening Deep Test of Articulation. The intra- and interexaminer reliability of the judges was assessed. Results suggested that, as a group, both experienced and inexperienced listeners were capable of making consistent decisions about children's responses on the Screening Deep Test of Articulation. It was determined, however, that group averages may not reflect accurately the consistency of all members of the group or accurately estimate the population average. Variances for both groups were high and significant differences were found between the variances of the groups. Confidence intervals were reported and their clinical implications discussed.
本研究考察了一组听众的平均意见或平均一致程度在多大程度上实际代表了该组所有成员判断的一致性。22名经验丰富和经验不足的评判员对6名存在多种发音错误的儿童在麦克唐纳深层发音筛查测试中的录音反应进行评分。评估了评判员的内部和评判员间信度。结果表明，总体而言，经验丰富和经验不足的听众都能够对儿童在深层发音筛查测试中的反应做出一致的判断。然而，研究确定，组均值可能无法准确反映该组所有成员的一致性，也无法准确估计总体均值。两组的方差都很高，且两组方差之间存在显著差异。报告了置信区间并讨论了其临床意义。
