In primary schools across India, teachers and pupils naturally mix English with regional languages in class, but the tests that follow are always in English alone. This mismatch matters because the vast majority of children in government schools come from underprivileged backgrounds with no English exposure at home and minimal literacy support. Current unilingual assessments likely underestimate what these children actually know, since they cannot draw on their full language resources. The project will test whether a "dynamic assessment" approach—where students who struggle receive prompts in their home languages—gives a fairer and more accurate picture of their abilities. Researchers will compare traditional static tests, English-only dynamic assessment, and multilingual dynamic assessment in Grade 4 classrooms in New Delhi and Guwahati, tracking learning over time. If successful, the findings could shift how schools across the Global South assess language and content learning. Rather than forcing children to work in a single language, assessment could mirror the multilingual reality of the classroom. This would make testing more equitable for millions of underprivileged learners and could inform education policy in other highly linguistically diverse countries where English, local languages, and national languages compete for space in the curriculum.
View original technical description
Research conducted with learners educated in or through a second language has shown that the use of regional or home languages in the classroom benefits both language and content learning. However, the embracement of a multilingual model of language teaching and learning does not currently map to a multilingual approach in assessing language and content, partly because the construct of multilingual assessment is yet to be adequately defined and operationalised. The present project intends to design, implement and evaluate the construct of multilingual assessment in primary government schools in India in which English is taught as a subject or as the medium of instruction (EMI). Children attending these schools grow up and live in a highly linguistically diverse society which embraces multilingualism in the individual speaker primarily in oral language usage with code-mixing/translanguaging as a natural mode of communication. The vast majority of learners in government schools come from underprivileged socioeconomic backgrounds associated with no or minimal home literacy support and lack of any input in English in spoken or written mode. Thus, code-mixing with English and the regional languages is common in teacher-talk and teacher-pupil interaction as a reflection of sociolinguistic practices and a way to scaffold language and content learning. School-based assessment, however, has always been unilingual. This project aims to promote continuity among teaching, learning and assessment through the active exploitation of learners' entire language repertoire. A step to this direction is the use of dynamic assessment (DA), which forms an integral part of students' learning process: students unable to provide a correct answer are supported with prompts that facilitate progression with the assessment tool in use and with follow-up assessments. DA thus offers a more fine-grained picture of students' content and language abilities than traditional, static forms of assessment, since it considers actual learning outcomes as well as learning potential. Furthermore, DA creates an opportunity to learn by making assessment equitable, particularly for underprivileged learners with poor home literacy support. The first objective is to quantify and qualify the use of code-mixing in pedagogical activities in English language classrooms using audio recordings. The second objective is to examine whether DA affects students' English language and content learning over time, and whether a multilingual vs. unilingual (English-only) DA differentially affects learning progress and outcomes. We will recruit three groups of children attending Grade 4 in government schools in two sites, New Delhi and Guwahati, where different regional languages and educational policies related to the promotion of regional language literacy have been implemented so far. The students will be assessed with a) traditional, static assessment b) English-only DA and c) multilingual DA following a longitudinal design and using tasks that will engage learners' listening and reading comprehension. The findings will have ground-breaking implications for the assessment of content and language learning outcomes, promoting a paradigm-shift in multilingual DA. This new model is informed by equity in education and a holistic view of learning, whereby learners are allowed to rely on the totality of their language resources. There is high potential for extending the findings to multilingual contexts outside India, particularly in the Global South, where language education policies seek to address conflicting demands of high linguistic diversity in the population, English as a global language and the need to promote local languages in the curriculum.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know