目的·评价项目研制的可用于轻度认知功能障碍筛查的电子化认知评估系统的信度和效度,构建机器学习法判定模型并评估筛查效果。方法·采用分层随机的方法在上海和河南农村的社区、老年护理院及专科门诊抽取55岁以上的符合标准...目的·评价项目研制的可用于轻度认知功能障碍筛查的电子化认知评估系统的信度和效度,构建机器学习法判定模型并评估筛查效果。方法·采用分层随机的方法在上海和河南农村的社区、老年护理院及专科门诊抽取55岁以上的符合标准的老年人,由经过严格培训、操作规范的调查员对研究对象进行蒙特利尔认知评估量表(Montreal Cognitive Assessment,MoCA)的现场测试。电子化认知评估系统信度评价采用内部一致性系数,效度评价采用因子分析;以MoCA评估结果作为标准,使用分类准确率和曲线下面积(area under curve,AUC)比较朴素贝叶斯、随机森林、Logistic回归和K-邻近4种机器学习算法的分类效果。结果·研究的359名对象中,年龄中位数为63岁,82.80%为中学及以下学历;根据MoCA评分,可能患有轻度认知功能障碍的有147名。电子化认知评估系统的Cronbach'sα为0.84,KMO为0.78,Bartlett's球形检验P<0.05,共提取13个公因子,累计方差贡献率为75.10%。最优朴素贝叶斯分类模型的分类准确率为88.05%,AUC为0.941。结论·该电子化认知评估系统具有良好的信度、效度及分类效果,利用朴素贝叶斯分类模型分类准确度较高。展开更多
This paper develops Davis technology acceptance model by adding two criteria of e-readiness assessment. For small and medium size enterprises, the paper evaluates the credit of developed model in studying the efficien...This paper develops Davis technology acceptance model by adding two criteria of e-readiness assessment. For small and medium size enterprises, the paper evaluates the credit of developed model in studying the efficient factors in internet technology acceptance, which is one of the most applicable communication tools in e-commerce. The added factors, taken from models and means of e-readiness assessment, are ＂existing information technology infrastructures＂ as well as ＂organization＇s training＂. The authors study 45 small and medium size enterprises in the city of Tehran, which are active in information technology products and services. Using Lisrel structural equation modelling, it is shown that there is a meaningful relationship between the added factors and the variables indicating internet network usage. Therefore, the better factors of e-readiness assessment lead to the higher acceptance level of electronic communication tools in e-commerce.展开更多
文摘目的·评价项目研制的可用于轻度认知功能障碍筛查的电子化认知评估系统的信度和效度,构建机器学习法判定模型并评估筛查效果。方法·采用分层随机的方法在上海和河南农村的社区、老年护理院及专科门诊抽取55岁以上的符合标准的老年人,由经过严格培训、操作规范的调查员对研究对象进行蒙特利尔认知评估量表(Montreal Cognitive Assessment,MoCA)的现场测试。电子化认知评估系统信度评价采用内部一致性系数,效度评价采用因子分析;以MoCA评估结果作为标准,使用分类准确率和曲线下面积(area under curve,AUC)比较朴素贝叶斯、随机森林、Logistic回归和K-邻近4种机器学习算法的分类效果。结果·研究的359名对象中,年龄中位数为63岁,82.80%为中学及以下学历;根据MoCA评分,可能患有轻度认知功能障碍的有147名。电子化认知评估系统的Cronbach'sα为0.84,KMO为0.78,Bartlett's球形检验P<0.05,共提取13个公因子,累计方差贡献率为75.10%。最优朴素贝叶斯分类模型的分类准确率为88.05%,AUC为0.941。结论·该电子化认知评估系统具有良好的信度、效度及分类效果,利用朴素贝叶斯分类模型分类准确度较高。
文摘This paper develops Davis technology acceptance model by adding two criteria of e-readiness assessment. For small and medium size enterprises, the paper evaluates the credit of developed model in studying the efficient factors in internet technology acceptance, which is one of the most applicable communication tools in e-commerce. The added factors, taken from models and means of e-readiness assessment, are ＂existing information technology infrastructures＂ as well as ＂organization＇s training＂. The authors study 45 small and medium size enterprises in the city of Tehran, which are active in information technology products and services. Using Lisrel structural equation modelling, it is shown that there is a meaningful relationship between the added factors and the variables indicating internet network usage. Therefore, the better factors of e-readiness assessment lead to the higher acceptance level of electronic communication tools in e-commerce.
