论文部分内容阅读
测量的实质就是比较。作为心理测量,考试是一种特殊的测量,它必须依赖于一系列前提假设,才能够使得它的比较具有意义。通过分析现有的考试理论和它们的前提假设,本文提出了考生群体分布一致性假设,使用HSK实际考试的数据,对于该假设进行了统计检验。基于对于这个假设的验证,本文设计了考生分层抽样的方法,获取了分布形态保持高度一致的考生样本。考生群体分布一致性假设的验证,可以验证和补充现有考试理论的一些假设,分布形态高度一致考生样本的获得,实际上就是一个稳定参照标准的获得,与之进行比较就能够获得有意义的测量结果。
The essence of measurement is comparison. As a measure of psychology, a test is a special measure that must rely on a set of assumptions that make it more meaningful. By analyzing the existing test theories and their assumptions, this paper presents the hypothesis of the consistency of candidates’ population distribution, and uses the data of HSK practical test to test the hypothesis. Based on the validation of this hypothesis, this paper designs a method of stratified sampling for candidates, and obtains the samples of candidates whose distribution patterns remain highly consistent. Candidate group distribution consistency hypothesis verification, you can validate and supplement some of the existing test theory assumptions, distribution patterns highly consistent sample of candidates, in fact, is a stable reference standard to obtain, compared with it to obtain meaningful Measurement results.