Abstract

The Controlled Oral Word Association Test (COWAT) is a widely utilized measure of phonemic fluency. However, two issues remain: (1) whether demographic, cognitive variables, or version of test administered predict performance; (2) if the test is predictive of Mild Cognitive Impairment (MCI). Recent studies report that item-level analyses such as lexical frequency may be more sensitive to early cognitive change. The purpose of this study was to examine the clinical utility of the COWAT, considering both total correct words and the lexical frequency. Sixty-seven healthy adults and thirty-seven adults with MCI completed neuropsychological testing. Mann–Whitney U tests were used to determine if there was a difference in COWAT performance between groups. Elastic net regression models were used to assess whether variance in total scores/lexical frequencies can be predicted by demographics, test version, or diagnosis; which cognitive tests explained the variance in performance; and how total scores and lexical frequencies compared with other cognitive tests in predicting diagnosis. Overall, individuals with MCI produced fewer and higher frequency words. The variance in total correct words or lexical frequency was not explained by demographics, test version, or diagnosis. Total correct words was a more important predictor of diagnosis than lexical frequency.

摘要

受控口头单词联想测试 (COWAT) 是一种广泛使用的音素流畅性测量方法。但是，仍然存在两个问题：(1) 人口统计学、认知变量或测试管理版本是否可以预测性能；(2) 如果该测试可预测轻度认知障碍 (MCI)。最近的研究报告说，词汇频率等项目级分析可能对早期认知变化更敏感。本研究的目的是检查 COWAT 的临床效用，同时考虑总正确词和词汇频率。67 名健康成年人和 37 名患有 MCI 的成年人完成了神经心理学测试。曼-惠特尼大学测试用于确定各组之间的 COWAT 性能是否存在差异。弹性网络回归模型用于评估总分/词汇频率的方差是否可以通过人口统计学、测试版本或诊断来预测；哪些认知测试解释了表现的差异；以及与预测诊断的其他认知测试相比，总分和词汇频率如何。总的来说，患有 MCI 的人产生的词汇更少但频率更高。人口统计学、测试版本或诊断无法解释总正确词或词汇频率的差异。总正确词是比词汇频率更重要的诊断预测指标。