摘要
采用锚题设计,选择年级1作为参照基准,通过蒙特卡洛模拟方法,考察锚题比例与年级离散度对垂直等值的影响.研究结果表明:与基准年级的距离影响垂直等值效果,越靠近基准年级,估计精度越好;从整体而言,垂直等值锚题比例设为30%,等值效果最好;垂直等值锚题比例的设定受年级离散度影响,2者存在交互作用,锚题比例设为"变"值更好.
The paper aimed at the effect of the ratio of anchor items to total test and the separation of grade distributions on the precision of vertical scaling. The paper used common-item design and chose grade 1 as base grade,by Monte Carlo simulation method simulated the different ratio of anchor items to total test and separation of grade distributions studied the influence of vertical scaling. The result shows that:( i) Selection of base grade affected the precision of vertical scaling. The closer to the base grade,the better the accuracy of the estimate is.( ii) When there is a 30% ratio of anchor items to total test,the precision of vertical scaling is best.( iii) The ratio of anchor items to total test is affect by the separation of grade distributions. There is an interaction between the two,and the ratio of anchor questions is set to " change".
作者
黎光明
梁正妍
LI Guangming;LIANG Zhengyan(School of Psychology,Center for Studies of Psychological Application,South China Normal University,Guangzhou Guangdong 510631,China)
出处
《江西师范大学学报(自然科学版)》
CAS
北大核心
2019年第1期52-58,共7页
Journal of Jiangxi Normal University(Natural Science Edition)
基金
国家自然科学基金(31470050)
教育部人文社会科学研究规划基金(18YJA190006)
广东省哲学社会科学"十三五"规划一般课题(GD17CXL01)
广州市哲学社会科学"十三五"规划一般课题(2017GZYB111)资助项目
广东省2015年度高等教育教学改革课题(粤教高函[2015]173号)
关键词
锚题比例
年级离散度
垂直等值
测验等值
项目反应理论
ratio of anchor items to total test
separation of grade distribution
vertical scaling
test equating
item response theory