The purpose of this study was to examine the impact of clinical review bias on the coding of grammaticality in child language. Seventy-four native-English students studying communication disorders and sciences made judgments about the grammaticality of 250 utterances presented in five language samples. Each language sample included grammatical, ungrammatical, and ambiguous utterances. Participants were randomly assigned to a blind or nonblind group. The nonblind group was presented with diagnostic information, whereas the blind group was not. We employed a generalized linear mixed model to examine the binary data. Our results suggest that both blind and nonblind participants were accurate in judging grammatical and ungrammatical utterances. However, nonblind participants were slightly more likely to judge ambiguous utterances as ungrammatical when the language sample identified the child as having a language impairment, suggesting that there was an effect of clinical review bias in this study. This effect, although statistically significant, was small.

中文翻译：

---

临床评价偏差对儿童语言语法的影响

本研究的目的是检查临床评论偏倚对儿童语言语法编码的影响。74 名学习交流障碍和科学的母语为英语的学生对 5 个语言样本中的 250 个话语的语法性做出了判断。每个语言样本都包括语法、不合语法和歧义的话语。参与者被随机分配到盲组或非盲组。非盲组获得了诊断信息，而盲组则没有。我们采用广义线性混合模型来检查二进制数据。我们的结果表明，盲人和非盲人在判断语法和非语法话语方面都是准确的。然而，当语言样本将儿童识别为有语言障碍时，非盲参与者更有可能将模棱两可的话语判断为不合语法，这表明本研究中存在临床审查偏倚的影响。这种影响虽然具有统计学意义，但很小。