Protest against inequitable implementation of language of instruction policies to resolve the learning crisis

• In Event: Educational Policy, Social Change and Equity

Proposal

Motivation
There is a growing body of evidence suggesting that teaching children in a language they understand leads to significantly better learning outcomes and appropriate grade progression (Crawford and Venegas, 2021; Spaull, 2016). The language of instruction used in an educational system is, therefore, one of the most impactful policy features that may drive learning outcomes.
Yet, in many low- and middle-income countries that were former colonies, the official language of instruction is still that of the colonizing country, despite the fact that many pupils and teachers may not be proficient in that language. It is estimated that between 4 and 5 out of every 10 children worldwide learn in a language other than their mother tongue (Crawford and Venegas, 2021; McBride et al., 2017). In other words, a large number of children globally, are expected to learn curriculum content while simultaneously learning the language in which it is being taught (Erling et al., 2017). In Uganda, a thematic curriculum has been in operation at primary level since 2007. The curriculum requires teaching around themes that are familiar to the learners and using predominant local language as the medium of instruction in the lower grades (Primary 1-3), transitioning to using English as a language of instruction from Primary 4 until the end of the 7-year cycle, while teaching English as a subject right from Primary 1. Despite this curriculum, many Ugandan children are not taught in their mother language due to language diversity and other socioeconomic factors.
Shifting towards local languages is a protest against inequitable policies and a vital step towards resolving the learning crisis. Our work explores the considerations necessary for Ugandan policymakers to establish a successful language of instruction policy implementation as a vital step towards resolving the challenge of inequitable learning outcomes. We take into account the linguistic landscape, past interventions, and the learning levels of pupils based on their language majority/minority status and school's language of instruction.
Research Questions
● What does the current linguistic landscape in Uganda look like, and how does it shape the implementation of the language of instruction policy?
● What are the learning outcomes of pupils attending schools that use local languages in Uganda?
● Do linguistic minorities in their communities achieve lower learning outcomes relative to their peers?
● How can a change in language of instruction implementation be approached in an equitable and inclusive manner, including the selection of a local language in each community?
Methods and Data
The Uwezo 2021 datasets, providing nationally representative learning outcomes data for pre-primary and primary-aged children, are used to answer these research questions. Descriptive statistics and regression analyses with appropriate sample weights ensure representative findings. Combined with existing literature on language of instruction policies, these insights offer unique perspectives on the current landscape and the challenges policymakers in Uganda and similar low- and middle-income countries might face during the transition to a mother tongue-based language of instruction policy.

Preliminary Findings
First, we confirm that Uganda is a linguistically diverse country. The Uwezo 2021 data reveals that linguistic diversity exists even within smaller geographic areas. Approximately 60.6% of areas have a linguistically homogeneous population, while the remaining areas exhibit varying degrees of linguistic diversity. For example, targeting Luganda, Lusoga, Lugbarati, Runyakitara, and Ateso as some of the widely spoken regional languages in Uganda, could reach 38% of all pupils in Uganda. However, 37% of pupils live in highly linguistically heterogeneous areas, making it challenging to implement a common language of instruction. In fact, 1 in 4 local areas in the Uwezo data have a significantly linguistically diverse population.
Learning outcomes vary based on the language of instruction and pupils' linguistic backgrounds. Pupils in English-only instruction outperform those in local language schools. Factors like pupil-teacher ratios, enrollment sizes, and early childhood education availability contribute to these disparities. Household wealth, language spoken at home, and learning outcomes are correlated. Pupils in wealthier households achieve higher learning outcomes, adding an equity lens to language of instruction selection. Lastly, our findings highlight the disparities faced by linguistic minority groups, who consistently exhibit lower learning levels compared to their peers from the dominant language group in their area. Policymakers need to consider the needs and learning outcomes of these minority groups to ensure equitable education provision.
Contribution
Existing studies have evaluated the causal impact of mother-tongue based instruction in low- and middle-income countries (Brunette et al., 2019; Chicoine, 2019; Kerwin and Thornton, 2021). High-level reports also highlight the benefits of mother tongue-based instruction at a national level, such as those by Crawford and Venegas (2021) and UNICEF (2016). However, there is a scarcity of literature exploring the implementation of mother tongue policies considering the specific linguistic landscape and learning levels of a particular country. Our work fills this gap by documenting how the Ugandan context, explored through Uwezo 2021 dataset, can inform the implementation of the language of instruction policy locally. Consequently, our findings suggest that policymakers in Uganda and beyond should take into account:
● The linguistic features of the target language: Certain languages pose greater challenges in developing programs. For instance, languages with complex grammatical systems or a large number of homophones can be more difficult for both learners and teachers.
● The availability of high-quality materials and resources: It is crucial to have high-quality teaching materials and resources accessible in the target languages. Given the socioeconomic disparities across linguistic groups, it is likely that different groups also have varying levels of access to educational inputs. Ugandan policymakers must consider this reality to develop equitable solutions wherever possible.
● The training and support of teachers: Teachers need to receive adequate training and support in effectively teaching in a mother tongue. This aspect plays a vital role in the success of mother tongue-based instruction interventions.
● The involvement of parents and communities: Parents and communities should actively participate in the implementation of mother tongue-based instruction interventions. Their support is valuable in assisting teachers and learners, and their explicit endorsement ensures the correct implementation of interventions.

Authors

• Daniel Rodriguez Segura, University of Virginia

• Mary Goretti Nakabugo, Uwezo Uganda