High-stakes tests as gatekeepers: Interrogating test architecture in Bangladesh

In this post, Maksud Ali, Obaidul Hamid & Ian Hardy from the University of Queensland, discuss their recent paper ‘Ritualisation of testing: problematising high-stakes English-language testing in Bangladesh‘, published in COMPARE, 2020 (vol 50, Issue 4).

High-stakes testing has become an inevitable part of Bangladeshi society. In this densely populated country, education and employment opportunities are limited and highly competitive. Therefore, tests are used as gatekeepers in both sectors. Despite this critical role which high-stakes tests play in making life-changing decisions for young people, the architecture of these tests is rarely questioned. Instead, tests are treated as social rituals which are performed and celebrated every year across the country, even as they create distinctive ‘winners’ and ‘losers’.

Students taking a high-stakes secondary test in Bangladesh (Financial Express, 2017)   Given their increasing significance in society, it is important to interrogate and understand how high-stakes tests are designed and what criteria guide their construction. In keeping with this goal, our study (Ali, Hamid & Hardy, 2020) examined the design of the Higher Secondary Certificate (HSC) English Paper 1 test. We analysed relevant test policies and interviewed test writers and moderators to understand how the test was aligned with the higher secondary level English curriculum, which emphasises human capital development with a focus on employment-relevant English language skills among students. See our paper .

The study revealed a significant gap between the curricular goals and the test architecture. The test and the curriculum had divergent goals— i.e., while the curriculum emphasises developing learners’ communication skills by including the skills of reading, writing, listening and speaking, the test did not have any provision for assessing listening and speaking skills. Importantly, the test does not articulate any explicit aims, which made it difficult to understand what the test intends to assess. Similarly, there is a lack of test specifications for test writers and moderators, who struggle to align the test with the curriculum. As a result, they often end up aligning the test with a problematic test model provided by the Bangladesh National Curriculum and Textbook Board (NCTB). The model paper is problematic because it does not fully represent the curriculum.

Problematic test policies have impacted on test practices. We reveal test writers and moderators developed “compliant dispositions” as their roles were limited to responding to the prescribed model. Test policy and practices are also influenced by contextual issues, including: (a) corruption, and; (b) urban-rural social, economic and education disparities. As our data reveal, speaking and listening were not introduced in the assessment as the authorities believed that testing these skills might encourage some teachers to make money by taking advantage of students’ difficulties in these areas. It is also believed that testing these skills would disadvantage students from poor socioeconomic backgrounds, who, unlike their counterparts in urban contexts, do not have access to sufficient learning opportunities and resources at home and at school.

These findings have clear implications for test users. Although test data are used across society for making life-changing decisions, the data may not be valid in terms of broader educational and curricular goals. In countries like Bangladesh, test practices are comparable to social rituals, which show limited concern about validity, reliability and the ethicality of tests. Construction and consumption of tests and test data are part of broader social practices which are not interrogated but celebrated as taken for granted social norms.

References

Ali, M. M., Hamid, M. O., & Hardy, I. (2020) Ritualisation of testing: problematising high-stakes English-language testing in Bangladesh. Compare: A Journal of Comparative and International Education, 50(4), 533-553, DOI: 10.1080/03057925.2018.1535890

Financial Express (2017 November 23). SSC, equivalent exams to begin Feb 1. The Financial Express. Retrieved from http://thefinancialexpress.com.bd/education/ssc-equivalent-exams-to-begin-feb-1-1511407958

Photo, published with permission.

About the Authors: Md. Maksud Ali is a PhD candidate at the School of Education, The University of Queensland, Australia. He is researching broader policy issues and practices of human capital development in English language education in Bangladesh. His research has been published in English Today, Asian EFL Journal, Compare: A Journal of Comparative and International Education, and Language Assessment Quarterly. Email: mdmaksud.ali@uq.edu.au

M. Obaidul Hamid is Senior Lecturer in TESOL Education at the University of Queensland, Australia. Previously he worked at the University of Dhaka, Bangladesh. His research focuses on the policy and practice of TESOL education in developing societies. For further details of his work, please see his website at this link: : https://researchers.uq.edu.au/researcher/1604

Ian Hardy is Associate Professor in Education at the School of Education, The University of Queensland. Dr Hardy’s work is informed by research into the relationship between education and society, particularly broader policy and political discourses, and educators’ responses to the socio-political contexts in which their work is undertaken. His forthcoming volume ‘School Reform in an Era of Standardization: Authentic Accountabilities’ will be published by Routledge in 2021. For further details of his work, please see his website at this link: https://researchers.uq.edu.au/researcher/990