摘要
现在的心理(或教育)测量学教科书一般认为评价一份测验的质量有两个标准:信度和效度。但是,实践发现,高信度和高效度的测验对某些学生群体也可能存在测验偏差,即冒犯或不公平地惩罚了某些学生群体。因此,测验编制者必须了解测验偏差的含义和测验偏差的数学模型,掌握测验偏差的估算与检验方法以及如何对测验偏差进行审查。
Reliability and validity were generally considered test quality standard. Actually, the test with good reliability and validity for certain groups of students may also exist test bias, which were offending or unfairly punishing some student groups. So the test bias and the meaning of test bias of the mathematical model should be understood. The test deviation estimation and test methods as well as the deviations review should be mastered.
出处
《韶关学院学报》
2013年第7期130-134,共5页
Journal of Shaoguan University