Ishigami Yoko, Eskes Gail A, Tyndall Amanda V, Longman R Stewart, Drogos Lauren L, Poulin Marc J
Department of Psychiatry, Faculty of Medicine, Dalhousie University, Halifax, Canada.
Department of Psychology, Faculty of Science, Dalhousie University, Halifax, Canada.
Exp Brain Res. 2016 Mar;234(3):815-27. doi: 10.1007/s00221-015-4493-4. Epub 2015 Dec 8.
The Attention Network Test (ANT) is a frequently used computer-based tool for measuring the three attention networks (alerting, orienting, and executive control). We examined the psychometric properties of performance on a variant of the ANT, the Attention Network Test-Interaction (ANT-I) in healthy older adults (N = 173; mean age = 65.4, SD = 6.5; obtained from the Brain in Motion Study, Tyndall et al. BMC Geriatr 13:21, 2013. doi: 10.1186/1471-2318-13-21) to evaluate its usefulness as a measurement tool in both aging and clinical research. In terms of test reliability, split-half correlation analyses showed that all network scores were significantly reliable, although the strength of the correlations varied across networks as seen before (r = 0.29, 0.70, and 0.68, for alerting, orienting, and executive networks, respectively, p's < 0.05). In terms of construct validity, ANOVAs confirmed that each network score was significant (18.3, 59.4, and 109.2 ms for the alerting, orienting, and executive networks, respectively, p's < 0.01) and that these scores were generally independent from each other. Importantly, for criterion validity, a series of hierarchical linear regressions showed that the executive network score, in addition to demographic information, was a significant predictor of performance on tests of conflict resolution as well as verbal memory and retrieval (β = -0.165 and -0.184, p's < 0.05, respectively). These results provide new information regarding the reliability and validity of ANT-I test performance in a healthy older adult population. The results provide insights into the psychometrics of the ANT-I and its potential utility in clinical research settings.
注意力网络测试（ANT）是一种常用的基于计算机的工具，用于测量三种注意力网络（警觉、定向和执行控制）。我们研究了健康老年人（N = 173；平均年龄 = 65.4，标准差 = 6.5；数据来自“运动中的大脑研究”，廷德尔等人，《BMC老年医学》13:21，2013年。doi: 10.1186/1471-2318-13-21）中注意力网络测试交互版（ANT-I）的表现的心理测量特性，以评估其在衰老和临床研究中作为测量工具的有用性。在测试信度方面，分半相关分析表明，所有网络得分都具有显著的信度，尽管各网络间相关性的强度如之前所见有所不同（警觉、定向和执行网络的相关系数分别为r = 0.29、0.70和0.68，p值均<0.05）。在结构效度方面，方差分析证实每个网络得分都具有显著性（警觉、定向和执行网络的得分分别为18.3、59.4和109.2毫秒，p值<0.01），且这些得分通常相互独立。重要的是，在效标效度方面，一系列分层线性回归表明，除人口统计学信息外，执行网络得分是冲突解决测试以及言语记忆和检索表现的显著预测指标（β分别为 -0.165和 -0.184，p值均<0.05）。这些结果提供了关于ANT-I测试在健康老年人群中表现的信度和效度的新信息。这些结果为ANT-I的心理测量学及其在临床研究环境中的潜在效用提供了见解。
