摘要
天花板和地板效应是增值评价中经常会涉及的问题。其不仅与测验本身的天花板和地板效应相联系,还与增值评价模型的选择有关。残差增分模型和学生成长百分等级模型是增值评价中常用的条件状态模型,其所估计的个体成长是与同等水平学术伙伴比较的相对成长,在一定程度上可以破解天花板和地板效应问题。基于两组考生的中、高考数据进行分析比较,残差增分模型和学生成长百分等级模型在解决天花板和地板效应问题上的有效性得到了验证。研究结果为研究者和教育评价工作者选择增值模型提供了参考。
Ceiling and floor effects are issues frequently concerned in value-added assessment.Ceiling and floor effects in value-added assessment are related not only to ceiling and floor effects in tests,but also to the application of value-added models.The residual gain model and student growth percentile model are conditional status models commonly employed in value-added assessment.The estimated individual growth in these two models represents normative growth compared to academically similar peers with comparable prior scores,thereby partially resolving the issues associated with ceiling and floor effects in value-added assessment.By analyzing Zhongkao and Gaokao data from two distinct student groups,this study provides empirical support for the efficacy of the residual gain model and student growth percentile models in addressing ceiling and floor effects.Assessment scholars and practitioners can refer to these findings when selecting appropriate value-added models.
作者
骆文淑
陈方丁
LUO Wenshu;CHEN Fangding(Foshan University,Foshan 528225,Guangdong,China;Education Examinations Authority of Guangdong Province,Guangzhou 510635,China)
出处
《中国考试》
北大核心
2024年第3期87-97,共11页
journal of China Examinations
基金
广东省2022年度教育科学规划课题(教育综合改革专项)“增值评价模型的理论和应用研究”(2022JKZG082)。
关键词
增值评价
增值模型
天花板效应
地板效应
value-added assessment
value-added model
ceiling effect
floor effect