摘要
应用概化理论对结构化面试的评分误差的控制问题进行了研究。结果表明:结构化面试评分能够较好地反映出被试的真实能力水平,评分具有较高的信度;在保证较高的面试评分信度(0.80)的情况下,建议将考官人数减少至9名,以提高结构化面试的经济性和效率性。
This study applies generalizability theory to control the assessing error in structural interview, and draws the following conclusions： The scores assessed by rater in the structural interview can reliably reflect the subject＇s true competency; Under the circumstance of high assessment reliability （0.80）, the study suggests to reduce the raters to 9 in order to economically and efficiently improve the structural interview.
出处
《西北师大学报（社会科学版）》
CSSCI
北大核心
2006年第4期62-65,共4页
Journal of Northwest Normal University(Social Sciences)
关键词
概化理论
结构化面试
评分误差
generalizability theory
structural interview
assessing error
