Lessons and key descriptive findings from a cross-national study on language of instruction transitions.

• In Event: Language of instruction transitions: factors that predict multilingual literacy acquisition and cross-linguistic skills transfer from four countries

Proposal

Despite the complexity of language in education policy issues and the salience of language acquisition for effective foundational skill acquisition, there is a lacuna of data and evidence related to key factors affecting language and literacy acquisition in LMICs. This multi-country study promises to make important contributions and offer a robust source of evidence.


In this initial presentation, we will introduce the study, Language of Instruction Transitions in Education Systems (LITES), situate it within the USAID-funded SHARE activity (Sustaining Holistic and Actionable Research in Education), describe an overview of the study design, partnership structure, and methodology employed, and present salient and noteworthy findings from key descriptive statistics.


A brief description of the aims of SHARE and research partnership model will first be presented. SHARE embraces the dual aims of advancing USAID’s education learning agendas through multi-country studies with the aim of robust localization and capacity strengthening of local research institutions. As such, the model is both novel and ambitious. A description of this approach and key lessons will briefly be shared.


Additionally, recognizing that the study offers many important contributions beyond the specific research questions posed, we wish to first shine a spotlight on key methodological decisions and share notable lessons and insights related to instrument design, sampling approach, and key descriptive data.


The study faced complex questions related to sampling, language selection, and feasibility. Testing students in multiple languages was regarded as essential and a major gap in the research and national assessments from development projects, but posed logistical challenges. The role of Kiswahili, as a national language in Kenya, alongside 40 local languages and English as the target L2, posed a particular methodological and logistical quandary.


Examples of key descriptive data across the four study contexts include the following:
Teacher attitudes towards multilingual education and local language instruction;
Teachers’ sense of efficacy for multilingual instruction;
Language use patterns in the classroom;
The availability of teaching and learning materials in L1 and L2;
Teachers’ level of L1/L2 proficiency;
Students’ decoding abilities in L1/L2
Students’ level of L1/L2 proficiency;
Students’ level of decoding/reading fluency in L1/L2;
The degree of language match between students and the languages of instruction used
Students’ reading comprehension in L1/L2, etc.


Finally, the study implemented novel approaches to measuring students reading comprehension, employing a traditional EGRA reading comprehension task, a modified version of this task (allowing students to “look back” at the text to mitigate the conflation of short-term working memory with reading comprehension), and a new measure of students’ reading comprehension using a 7-10 item task with a short text and multiple choice picture matching with graduated difficulty.


A brief presentation of some of these key statistics, key methodological choices, and the successes, challenges, and learning from these methods, will be shared prior to the presentation of the results for each of the three major research questions.

Authors

• Anthony Ssebagereka, ResilientAfrica Network Makerere University School of Public Health

• Pooja Nakamura, American Institutes for Research

• TJ D'Agostino, University of Notre Dame