Department of Psychology and Brain Research New Zealand, University of Otago.
Department of Psychology, University of Otago.
Psychol Assess. 2018 Apr;30(4):539-549. doi: 10.1037/pas0000503. Epub 2017 May 29.
Many longitudinal studies entail repeated computerized cognitive testing. However, few of the measures frequently employed for assessing cognitive changes have been examined for within- and between-day retest effects over multiple time points. To address gaps in the repeated cognitive assessment literature, in the current study 20 male undergraduate students completed a computerized cognitive battery 6 times over 3 testing days, each 1 week apart. Performance shifts and test-retest reliability (TRR) varied for different measured variables. Reaction times (RTs) on tasks designed to measure inhibition, switching, and selective attention generally showed good TRR, and few or only gradual improvements over time, suggesting that these measures are suitable for use in repeated testing, and that the inclusion of brief familiarization periods sufficed to resolve initial practice effects. Similar findings were observed for both accuracy and speed on a 2-back working memory task. In contrast, RTs on a basic visuomotor task showed worsening performance across time (likely reflecting boredom effects) and poor TRR, suggesting unsuitability for measuring change in cognitive performance. Cost scores designed to assess inhibition, switching, and selective attention showed poor TRR, as did scoring metrics from visuospatial short-term and working memory tasks (Corsi block paradigm). This low reliability suggests that these scores may not be suitable for use in repeated assessment, and calls into question their validity as measures of the specified cognitive functions. Researchers planning studies entailing computerized assessment of cognitive functioning in healthy young adults should consider the current results when selecting suitable cognitive tasks and measures. (PsycINFO Database Record
许多纵向研究都需要重复进行计算机认知测试。然而，评估认知变化的常用指标中，很少有在多个时间点上进行过日内和日间重测效应的检验。为了弥补重复认知评估文献中的空白，在当前研究中，20 名男性大学生在 3 个测试日中，每个测试日相隔 1 周，共完成了 6 次计算机认知测试。不同测量变量的表现变化和测试-重测信度（TRR）各不相同。旨在测量抑制、转换和选择性注意的任务的反应时间（RT）通常具有良好的 TRR，并且随着时间的推移几乎没有或只有逐渐的提高，这表明这些测量方法适合在重复测试中使用，并且包含简短的熟悉期足以消除初始练习效应。在 2 位工作记忆任务的准确性和速度上也观察到了类似的发现。相比之下，基本视觉运动任务的 RT 随着时间的推移而变差（可能反映出无聊效应），并且 TRR 较差，这表明其不适合测量认知表现的变化。旨在评估抑制、转换和选择性注意的成本得分以及视空间短期和工作记忆任务（Corsi 块范式）的评分指标的 TRR 也较差。这种低可靠性表明，这些分数可能不适合用于重复评估，并且质疑它们作为指定认知功能的测量指标的有效性。计划对健康年轻成年人进行计算机认知功能评估的研究人员在选择合适的认知任务和指标时，应考虑当前的研究结果。
