The Effects of an Experimental Course Using Data-Driven Learning Approach in Chinese as a Second Language

The Data-Driven Learning (DDL) approach advocates for learners to transition from passive recipients of knowledge to active researchers by using typical, large-scale, and context-rich target language inputs to drive a bottom-up learning process (Johns, 1990). Given the scarcity of empirical research on applying DDL in Chinese as a Second Language (CSL) classrooms, this study designed a teaching experiment for confusable words to explore the effects and learners' attitudes towards the approach. Five advanced-level CSL learners with different native language backgrounds were recruited, and taught ten sets of confusable words over a five-week period. The first five lessons used an indirect DDL method, while the latter five used a direct DDL method with Sketch Engine. Prior to the course, a pre-test was conducted, and after completion, a questionnaire, post-test, and interviews were administered. The post-test showed an average improvement of 24% over the pre-test, and the Wilcoxon signed rank test indicated a statistically significant improvement in applying DDL to learning confusable words. Furthermore, learners had a positive attitude towards the course, generally favoring learning vocabulary through collocations, and preferred to observe pre-selected concordance lines under the guidance of the instructor.