To Err is Human: "Abnormal" Neuropsychological Scores and Variability are Common in Healthy Adults
To Err is Human: "Abnormal" Neuropsychological Scores and Variability are Common in Healthy Adults
复制标题
DOI：
10.1093/arclin/acn001
复制
发表时间：
2009-02-01
影响因子：
2.6
通讯作者：
Brooks, Brian L.
中科院分区：
文献类型：
--
作者：
Binder, Laurence M.;Iverson, Grant L.;Brooks, Brian L.
规范性研究的变异性能由健康成年人的神经心理电池进行审查。关于测试分数分散，规范的参与者往往有很大的差异之间的最好和最差的分数。当“异常”被定义为一个分数超过一个标准差低于平均值，在测试电池至少20个措施，绝大多数的正常参与者有一个或多个异常。将样本限制在高于平均智商或教育水平的参与者中，并使用更保守的异常定义，例如低于平均值的两个标准差，并不能消除异常分数的存在。我们的结论是，在电池中的神经心理测试的某些比例的异常性能是心理测量正常的。异常并不一定意味着获得性脑功能障碍的存在，因为低评分和大的个体内变异性往往是健康成年人的特征。我们建议测试组合开发人员提供关于正常样本变异量的数据，并提供具有假阳性率的基本率表，可在临床上用于解释测试性能。
Normative studies of variability in performance by healthy adults on neuropsychological batteries are reviewed. Regarding test score scatter, normative participants often have large discrepancies between best and worst scores. When "abnormality" was defined as a score more than one standard deviation below the mean, in test batteries with at least 20 measures, the great majority of normative participants had one or more abnormalities. Restricting samples to participants with above average IQ or educational levels and using more conservative definitions of abnormality, such as two standard deviations below the mean did not eliminate the presence of abnormal scores. We conclude that abnormal performance on some proportion of neuropsychological tests in a battery is psychometrically normal. Abnormalities do not necessarily signify the presence of acquired brain dysfunction because low scores and large intraindividual variability often are characteristic of healthy adults. We recommend that test battery developers provide data on the amount of variability in normal samples and also provide base rate tables with false positive rates that can be used clinically when interpreting test performance.