a Cognitive Science Research Center , University of Oklahoma , Norman , OK , USA.
Clin Neuropsychol. 2018 Apr;32(3):479-494. doi: 10.1080/13854046.2017.1368716. Epub 2017 Aug 23.
As the use of computer-based neurocognitive assessment is rapidly expanding, the need to systematically study and document key psychometric properties of these measures has become increasingly more salient. To meet this aim, this study examined test-retest reliability and practice effects for the Automated Neuropsychological Assessment Metrics General Neuropsychological Screening battery (ANAM GNS) in a sample of 94 community dwelling adults.
ANAM GNS was administered and then repeated (alternate form) after 30 days. Test-retest reliability, practice effects, and the standard error of measurement were calculated. Using these estimates, reliable change indices were calculated to determine degree of performance change needed to exceed chance and measurement error (with 90% confidence interval).
The test-retest reliability for the ANAM composite score was .91. Performance significantly improved upon retest, but the effect size was small consistent with minimal practice effects. The threshold indicating change beyond chance or measurement error with 90% certainty was .9 (z-score).
Findings suggest that the ANAM GNS has excellent test-retest reliability upon retest at 30 days. Small practice effects can be expected. Change greater than .9 standard deviations in the ANAM composite score is likely to represent meaningful clinical change. This paper presents initial psychometric data from the ANAM GNS and supports its use as a reliable measure of cognition.
随着基于计算机的神经认知评估的应用迅速扩展，系统地研究和记录这些测量方法的关键心理计量学特性的需求变得越来越重要。为了达到这一目的，本研究在 94 名社区居住的成年人样本中，检查了自动化神经心理评估指标通用神经心理筛选测试（ANAM GNS）的重测信度和练习效应。
在 30 天后，对 ANAM GNS 进行了测试并重复（交替形式）。计算了重测信度、练习效应和测量误差的标准误差。使用这些估计值，计算可靠变化指数以确定需要超出机会和测量误差的性能变化程度（置信区间为 90%）。
ANAM 综合评分的重测信度为.91。在重新测试时，表现明显提高，但效果大小与最小的练习效应一致。表示有 90%把握改变超过机会或测量误差的阈值为.9（Z 分数）。
研究结果表明，ANAM GNS 在 30 天后的重测中具有极好的重测信度。可以预期会有较小的练习效应。在 ANAM 综合评分中，大于.9 个标准差的变化很可能代表有意义的临床变化。本文提供了来自 ANAM GNS 的初步心理计量学数据，并支持其作为可靠认知测量工具的使用。
