| Summary: | 博士 === 國立臺中教育大學 === 教育測驗統計研究所 === 101 === In the era of globalization, the trend towards learning Chinese as a foreign language (CFL) has become increasingly popular worldwide. The increasing demand in learning CFL has raised the profile of the Chinese proficiency test (CPT). There are four major CPT focuses on this including Hanyu Shuiping Kaoshi (HSK), Test of Chinese as a Foreign Language (TOCFL), Scholastic Assessment Test (SAT) Subject Test in Chinese with Listening, and Advance Placement (AP) Chinese Language and Culture exams. However, the majority of these tests are administered by the traditional paper and pencil tests format. Although there are many studies about developing the tools for learning CFL, the construction of computerized test for CFL is hard to find on the literatures. The aims of the present study are: adopting the Common European Framework of Reference (CEFR) for item development; providing a framework by using item response theory (IRT) as the scoring method; constructing computerized adaptive testing (CAT) system.
This study will analyze in depth the inadequacy of current CPT’s utilizing the common European framework of reference (CEFR) for language learning, teaching, and assessment to develop a set of reliability and validity standards for A level listening and reading CPT and CAT system. The data will be analyzed by applying IRT three-parameter logistic (3PL) model. One thousand five hundred and seventy-six participants recruited from Grace Christian Collage in Philippine were administered with Chinese listening and reading tests via CBT in September, 2010. In addition, the effectiveness of applying CAT among the three estimating methods, namely maximum likelihood estimation (MLE), expected a posteriori (EAP), and maximum a posteriori (MAP) will be investigated.
|
|---|
