As English continues to evolve as a global language, language proficiency tests face an important challenge: how can they include a variety of English accents while remaining both accurate and fair?
This study explores how exposure to different English accent varieties affects test-takers’ performance on Aptis listening tasks, as well as their attitudes toward equality, diversity and inclusion (EDI) in language testing.
A total of 120 intermediate to high-intermediate English learners from Belgian French, Arabic, and Singaporean backgrounds participated. They completed three versions of listening tasks modeled on Aptis, with each version featuring a different accent variety:
- British English
- shared-L1 English (matching participants’ first language)
- non-shared L2 English (an additional English accent of participants’ choice).
Before completing the tasks, participants also responded to a survey about their attitudes toward including diverse global English accents in a high-stakes listening test.
Key highlights
- Performance across accents: Test-takers’ scores did not differ significantly across the three accent varieties, although performance was slightly higher with British English.
- Attitudes toward accents: Survey responses revealed two main themes - French and Singaporean participants were generally more supportive of including diverse accents, while Arabic participants expressed lower acceptance of accent diversity.
- Attitudes versus performance: The relationship between participants’ attitudes and their test performance was weak and inconsistent.
These findings suggest that using intelligible Global English accents in listening assessments is unlikely to harm test performance. This supports calls for greater accent diversity in language testing, promoting more inclusive and representative assessments.
Citation
Kang, O. & Kostromitina, M. (2025). Equality, diversity and inclusion: Different varieties of accents in Aptis English listening tests. British Council. https://doi.org/10.57884/QV3J-5H54