REFINING AND VALIDATING PAUL NATION’S VOCABULARY SIZE TEST FOR A-LEVEL CANDIDATES IN PAKISTAN: PSYCHOMETRIC AND PREDICTIVE EVIDENCE

https://doi.org/10.5281/zenodo.18240745

Abstract

Vocabulary assessment is essential for evaluating students’ reading, writing, listening, and speaking skills. The Vocabulary Size Test (VST), developed by Nation and Beglar (2007) and commonly known as Paul Nation’s Vocabulary Size Test, is widely used to measure written receptive vocabulary. This study aimed to refine the VST for A-level candidates in Pakistan using a quantitative research design grounded in Classical Test Theory (CTT). Data were collected from 333 students across different institutes using convenience sampling. Item analysis revealed that 50% of the items fell within the acceptable facility value range (0.30–0.70), while 1% were too easy and 49% were too difficult, indicating that half of the items were moderately difficult and effective in discriminating among learners. Discrimination analysis further reduced the pool to 32 items with discrimination indices ≥ 0.30, most of which were drawn from higher vocabulary levels, demonstrating stronger performance in advanced lexical areas. Reliability analysis of the refined test yielded a Cronbach’s alpha of .763, exceeding the commonly accepted minimum threshold of 0.70 and indicating adequate internal consistency. Corrected item–total correlations were predominantly positive and of sufficient magnitude, confirming that the retained items showed adequate discrimination and contributed meaningfully to the overall scale while preserving content coverage across different lexical levels. Predictive validity was examined using simple linear regression, which showed that VST scores significantly predicted English achievement at the A-level, accounting for approximately 18% of the variance in achievement. Overall, the refined VST provided reliable and valid measurement of receptive vocabulary knowledge among A-level learners in Pakistan. The refined VST provides future researchers working with A-level students with a psychometrically sound measure of receptive vocabulary knowledge; in addition, it offers a stable empirical basis for subsequent model building and predictive analyses involving English achievement. In addition to its research utility, the refined VST can support diagnostic profiling of receptive vocabulary knowledge and inform preliminary placement decisions for A-level learners when used alongside complementary assessment evidence.