Enhancing predictive validity in the English placement test: evidence and insights from a Thai public university
摘要
In Thai higher education, first-year university admissions rely on multiple English test scores from both national and international examinations. Because these tests differ in purpose, constructs, and scoring criteria, administrators and teachers often have difficulty placing students into appropriate English courses, especially when all students are required to begin with the same foundation English course. To address this challenge, the Chulalongkorn University Language Institute (CULI) developed an English Placement Test (EPT) to classify students according to their actual proficiency and to support differentiated instruction. The test consists of 80 multiple-choice items (20 each on listening, vocabulary, language use, and reading) plus a one-paragraph writing task on a familiar or background-related topic. This study examines the predictive validity of the EPT by analyzing its correlation with achievement in an English foundation course. Data from 303 first-year health science EFL students were analyzed using Pearson product-moment correlations and regression analysis. Results showed moderate and significant correlations, indicating that the EPT effectively predicts course outcomes. The EPT was particularly useful in identifying low-proficiency students who required extra support, while mid- and high-proficiency groups generally performed well and met course expectations. Its diagnostic writing component also provided formative feedback, enabling teachers to better tailor instruction. The results highlight the value of placement testing for student grouping and pedagogical decision-making.