Show Less
Restricted access

Language Proficiency Testing for Chinese as a Foreign Language

An Argument-Based Approach for Validating the Hanyu Shuiping Kaoshi (HSK)


Florian Meyer

How did the (old) Hanyu Shuiping Kaoshi (HSK) assess the Chinese proficiency of non-native speakers of Chinese? What inferences can be derived from HSK test taker scores, especially from Western test takers? How difficult is it to learn Chinese according to the HSK? Thirty years of research have been synthesized into an argument-based approach for validating the most widespread test for Chinese as a foreign language. In addition, the author has analyzed the scores of a sample of over 250 German test takers in order to investigate how many hours German natives needed on average to study for reaching a specific HSK level. This work also extensively discusses validation theory for psychological testing, and it demonstrates how to put an argument-based approach to validation into practice.
Show Summary Details
Restricted access

4 An argument-based validation of the HSK


Chinese research clearly reveals that validation is a complex and continuing process. It is natural that in such a long-lasting process some ideas or procedures have to be modified or even refuted. So, one goal of this chapter is partly to reveal when and why important changes happened in the development and use of the old HSK, so the historical context also has to be taken into consideration. The main goal of this chapter is on the one hand to identify any inferential or causal chain gaps that may exist, and on the other hand to demonstrate where HSK research provides strong evidence supporting these inferences. Interestingly, Chinese HSK researchers have already recognized the idea that test validation elements have to be linked together into a systematic process. Guō Shùjūn and Gōng Wèi (1989) explain:

An examination is a systematic procedure. It includes setting up items, test administration, reviewing and annotating scores, composing scores, and interpreting/explaining scores, in addition to a series of other links. If one wants to make the examination accurate and reliable, one must strictly control all possibly occurring errors in every single link. (Guō Shùjūn and Gōng Wèi, 1989, p. 195)


You are not authenticated to view the full text of this chapter or article.

This site requires a subscription or purchase to access the full text of books or journals.

Do you have any questions? Contact us.

Or login to access all content.