Policy–Test Alignment in High-Stakes Assessment: Evaluating Construct Representation in China’s NMET Under the 2017 Curriculum Standards

Abstract

This study examines policy–test alignment in China’s National Matriculation English Test (NMET) by evaluating how assessment evidence represents the competencies articulated in the Senior High School English Curriculum Standards (ECS 2017). Recent curriculum reforms emphasize competencyoriented English education, yet the extent to which these policy goals are reflected in high-stakes examinations remains unclear. The analysis uses Kane’s Interpretation/Use Argument (IUA) to examine construct representation across NMET task domains, rather than treating alignment as surface correspondence between policy statements and test formats. The documentary analysis focused on ECS 2017 and January and June 2025 NMET papers used in Zhejiang Province, where nationally standardized papers are locally administered. Alignment was examined through expert judgment involving five specialists (κ = .366) and Webb’s alignment indices. The findings indicate that language competence is systematically represented across NMET sections, providing stronger construct representation evidence within this domain. In contrast, cultural awareness and thinking quality are represented selectively and unevenly, which limits the claims that can be made from NMET evidence. Learning ability is not instantiated in NMET task demands. Analysis of construct representation suggests a coherent assessment design but does not offset gaps in competency coverage. This study concludes that NMET evidence should be interpreted in domain-specific terms. It provides stronger evidence for some ECS 2017 competencies than for others, and it cannot support broad claims about the full curriculum framework.

Keywords

Full text

How to cite