基于多级评分的多维计算机化自适应测验技术及其应用研究

Multidimensional computerized adaptive testing (MCAT) is a combination of computerized adaptive testing (CAT) and multidimensional item response theory (MIRT) which shows great potential to obtain diagnostic information from examinees’ item responses accurately and efficiently, with its goal of extracting as much information as possible about the multiple abilities required and facilitating to help teaching and learning..Polytomously-scored items have been broadly used in a variety of tests for their advantages of providing more information and testing complicated abilities and skills. It is promising to develop MCAT adopting polytomously-scored items (PMCAT). In this project, all the research work are supposed to do in the PMCAT area and the research mode incorporating psychometric, statistic, modelling and mathematical validation are employed. The purposes of this project are the following..Firstly, it attempts to propose some algorithms to select polytomously or dichotomously scored items in PMCAT, which could be used in more test contexts..Secondly, it attempts to investigate the methods of item exposure controlling and to develop some item selection methods that unifies the demands of item estimation precision and item exposure uniformity..Thirdly, it tried to realize the diagnosis assessment of psychological diseases through PMCAT. Which is an important and valuable work in real practice. One is that the research could validate the feasibility and reasonability of PMCAT in real work, another is that the assessment on psychological diseases via PMCAT could be performed efficiently and accurately, and provide much more information which might be more useful to the individual. Thus the mode might be also work in other area. . Above all, this project is to fill the gaps of MCAT in practice work, and do some attempts on PMCAT to make some more accurate and efficient and broad its application fields.

多维计算机化自适应测验(MCAT）将多维项目反应理论（MIRT）与计算机化自适应测验(CAT)相结合，不仅能实现因人实测、提高测验效率，还能同时估计被试在测验各维度的能力水平，诊断个体的认知优势与不足，为个体的因材施教及有针对性地开展补救提供服务。本项目拟对当前具有重要理论及应用价值的基于多级评分的多维计算机化自适应测验(PMCAT)技术开展研究，采用心理测量、统计建模、数学论证多学科交叉研究范式：一方面开发出可以处理多级评分项目的PMCAT选题策略，以弥补传统二级评分MCAT的不足；另一方面，针对PMCAT的曝光控制方法进行研究，并开发出兼顾参数估计精度与项目曝光均匀性的选题策略；最后将PMCAT技术应用于心理障碍评估中，一方面在实践中验证PMCAT的可行性与合理性，另一方面为心理障碍评估实现高效准确快速地对被试诊断分类、并提供更为丰富和更具诊断价值的信息提供新的技术支持。

多维计算机化自适应测验(MCAT）将多维项目反应理论（MIRT）与计算机化自适应测验(CAT)相结合，不仅能实现因人实测、提高测验效率，还能同时估计被试在测验各维度的能力水平，诊断个体的认知优势与不足，为个体的因材施教及有针对性地开展补救提供服务。本项目对当前具有重要理论及应用价值的基于多级评分的多维计算机化自适应测验(PMCAT)技术开展研究，采用心理测量、统计建模、数学论证多学科交叉研究范式：一方面开发出可以处理多级评分项目的PMCAT选题策略，弥补了传统二级评分MCAT的不足；另一方面，针对PMCAT的曝光控制方法进行研究，并开发出兼顾参数估计精度与项目曝光均匀性的选题策略；最后将PMCAT技术应用于心理障碍评估中，一方面在实践中验证PMCAT的可行性与合理性，另一方面为心理障碍评估实现高效准确快速地对被试诊断分类、并提供更为丰富和更具诊断价值的信息提供新的技术支持。. 本项目共发表SCI/SSCI以及CSSCI期刊论文28篇（具体参见文末的附录），其中SCI/SSCI期刊论文14篇，《心理学报》4篇，《心理科学》10篇；发表的学术论文获得江西省教育科学优秀成果奖一等奖和三等奖各1项；出版学术著作《认知诊断理论》一部分在国际学术会议（如 NCME）上发表并报告学术论文 2篇；国内学术会议（如中国心理学大会）发表并报告学术论文 10 余篇。培养研究生17名，其中博士研究生3人（1人已毕业），硕士研究生14人（其中9人已毕业）。近四的研究中，项目主持人及项目组成员较好地完成了项目原定任务，另有一些研究突破了原申报书原定计划。