Examining the benefit of L2 language proficiency on academic performance using Bayesian logistic modeling

Area of Science:

• Educational Assessment
• Second Language Acquisition

Background:

• The Hanyu Shuiping Kaoshi (HSK) is a mandatory L2 proficiency test for international graduate students in China.
• The predictive validity of HSK scores for first-year graduate academic performance is under-researched.

Purpose of the Study:

• To investigate the relationship between HSK scores and first-year graduate academic performance.
• To assess the predictive power of HSK scores and undergraduate academic records for graduate success.

Main Methods:

• A Bayesian logistic regression model was developed using data from 666 graduate students in southern China.
• HSK scores (listening, reading, writing) and undergraduate GPAs were used as predictors.
• First-year graduate performance was binary (>=85% or <85%).

Main Results:

• Statistically significant correlations were found among HSK test components.
• The Bayesian model demonstrated strong predictive accuracy for achieving a B+ or higher.
• The model effectively integrated L2 proficiency and prior academic achievement.

Conclusions:

• HSK scores and undergraduate academic achievement are significant predictors of international students' success in Chinese graduate programs.
• L2 proficiency, measured by HSK, plays a crucial role in academic outcomes.
• Universities can utilize HSK scores as an early indicator of potential graduate academic performance.