摘要
大数据分析是一种以现代信息技术为基础的高级分析方法,其主要特征是数据的数量巨大,数据的结构、来源和形式复杂多样,数据的产生和获取过程常与实际工作相伴等。在简要探讨大数据分析的六类技术和八大服务企业、考察大数据分析技术在SAT-I常模研发方面的应用历史后,发现仅仅通过大数据分析很有可能导致常模错误或常模漂移等问题。为此,建议大数据分析技术必须与测评理论等同时使用,以保障测验常模的质量。
Big data analysis is an advanced analytic method based on modern information technology. The main characteristics of big data are high volume,high velocity and high variety. This paper examined six big data analytic techniques and eight big data analysis services providers,as well as historical applications of big data in the norming of SAT -I. The results indicate that big data analysis may lead to an inadequate test norm or a scale drafting problem without the use of testing theories. It is recommended that big data analysis methods and test theories should be used simultaneously to obtain a high quality test norm.
出处
《教育测量与评价》
2018年第2期5-11,共7页
Educational Measurement and Evaluation
关键词
大数据
常模
测量理论
big data, norming,testing theories
