Infra-Data: Tapping into the (In)Visible Reservoir of Educational Evidence from the Global South

In this blogpost, Hamza Sarfraz and Zain Ul Abidin introduce and explore the concept of ‘infra-data’ in education and how it can be utilized to develop inclusive, equitable, and contextually relevant educational policies and interventions in the Global South and address the question of ‘missing data’ for SDG4.

The Context

As the deadline for the Sustainable Development Goals (SDGs) looms ever closer, the recurring question of ‘missing data’ in education has emerged and re-emerged. Highlighted in NORRAG’s project, this ‘missing data’ is not merely an absence of evidence but a reflection of systemic blind spots in global education metrics and data regimes. But this missing data has real-life ramifications for how education policy and service-provision are conceptualized and implemented in the Global South.

SDG 4—with its promise of inclusive and equitable quality education—presumes to emerge from a shared universal epistemology in how its metrics are created, a notion that has been problematized elsewhere. In this blog, we look at how obsession with globally comparable metrics for SDG4 leads to ignorance, if not deliberate exclusion, of localized pedagogical practices and the wealth of lessons they provide. Beneath the officialized surface of these widely accepted metrics lies a vast sea of data, overlooked but ripe with information. This unseen data, which we term ‘infra-data’, emerges from successful educational interventions in the Global South and constitutes a constantly replenished reservoir of evidence on what works in education.

Drawing inspiration from James C. Scott’s concept of hidden transcripts and infrapolitics—in the sense of existing but not reducible to palpable conceptualizations—the notion of ‘infra-data’ looks at insights from those localized educational interventions, reflecting alternative knowledge systems that continuously negotiate with the all-encompassing influence of modern educational paradigms and broader modernity. This data, often beyond the public transcript, is a reservoir of untapped potential, offering a fresh perspective on learning, schooling, and teaching.

The idea of ‘infra-data’ also resonates with the concept of ‘abyssal lines’ as proposed by Boaventura de Sousa Santos, which delineates between the realms of “this side” of the line (where dominant knowledge systems reside) and “the other side” (which remains invisible and unaccounted for). Aníbal Quijano’s ‘coloniality of knowledge’ further underscores the systemic sidelining of localized and indigenous knowledge systems, emphasizing the need to recognize and value the rich data emerging from the peripheries.

In this article, we explore the concept of infra-data, its potential to reshape our understanding of educational outcomes, and its implications for policy, planning, and research in the Global South. We present cases from Pakistan—with 23 million ‘out-of-school’ children—and argue that, by acknowledging and harnessing infra-data, we can pave the way for more inclusive, equitable, and contextually relevant educational solutions, while also ensuring that the dream of SDGs is achieved before we run out of time.

Successful Educational Interventions – The Hidden Layer of Infra-Data

Across the Global South, particularly in Pakistan, the emergence of a variety of educational interventions, ranging from pedagogical innovations to edtech solutions, offer a wealth of ‘infra-data’. This data, deeply rooted in local contexts, encapsulates indigenous practices, pedagogies, and innovations that often remain overshadowed in global discussions.

Take, for instance, the Siyani Sahelian program, a second chance for vulnerable adolescent girls. Through its accelerated learning approach tailored to girls from a specific region from South Punjab, Pakistan, it has catered around 50,000 beneficiaries, of which it has successfully mainstreamed over 13,000 out-of-school girls into formal education. It has also developed its tools and learning trajectories utilizing data collected from the learners themselves.

Similarly, the Broad Class – Listen to Learn program by POWER99 Foundation harnesses the power of Interactive Radio Instruction, impacting over 200,000 beneficiaries across Pakistan, focusing on holistic development and improved teacher capacities. Alif Laila, which started as a mobile library, annually reaches tens of thousands of students, and has access to data on learning, teaching, and reading from all over Pakistan.

On the EdTech side, initiatives such as Mobile Taleem (200 schools) and Taleemabad (1 million+ children), showcase the potential of eLearning in diverse settings, from urban slums to refugee camps. These interventions leverage their own data to iteratively tweak their approach.

While many of these interventions have garnered global recognition from organizations like HundrED and WISE, their data often remains underutilized in the broader educational discourse. Beyond these, countless smaller yet equally impactful interventions exist, their data untapped, waiting to be integrated into a more inclusive and comprehensive understanding of education in the Global South.

The Potentialities of Infra-Data

With the context of international educational policy, ‘infra-data’ brings together a blend of diverse data strands. It encompasses a spectrum of data types, from learning outcomes, demographics, spatial and temporal patterns, to socio-cultural nuances, infrastructural elements, and beyond. The possibilities for deploying this data are endless.

1. Education Policy and Governance:

• Bridging Data Gaps by filling the chasm between ‘missing data’ and the available data on the learning situation across the Global South.
• Refining Global Metrics and SDGs by furnishing them with the depth of localized insights, ensuring that they resonate with the multifaceted realities of the billions of people.
• Enabling a multi-sectoral understanding of education and other critical sectors like public health, nutrition, gender, and inclusion, leading to integrated development strategies.

2. Service-Delivery:

• Empirical Richness with diverse data types, leading to innovative research methodologies and offering deeper insights into education patterns and practices.
• Challenging Algorithmic Biases in EdTech by providing the necessary training data with a more holistic basis.
• Providing Evidence from successful educational interventions of practices that can be adapted and scaled across diverse contexts.

3. Theory and Praxis:

• Broadening Learner Categories by including learners outside traditional schooling systems.
• Challenging Dominant Epistemes by emerging from localized practices and knowledge systems, advocating for a more pluralistic understanding of education.
• Decolonizing Knowledge—drawing inspiration from Aníbal Quijano’s ‘coloniality of power’—by emphasizing and foregrounding non-Western knowledge systems and lessons learned over decades of educational practice.

Infra-data, in its essence, is a multilayered amalgamation of diverse data types, capturing the multifaceted realities of education in the Global South.

The (Non)Existence of Infra-Data in Educational Discourse

The sidelining of infra-data in educational discourse is a nuanced issue, deeply entrenched in systemic biases, geopolitics, governance priorities, and dominant epistemologies. Its marginalization is not merely incidental but reflects broader policy debates and historical legacies, and thus leads to roadblocks in utilizing it.

1. Public Transcripts: Drawing from Scott’s idea, the official narratives or ‘public transcripts’ often exclude data that doesn’t fit official governing needs and narratives. Infra-data, being beyond these public transcripts, remains unrecognized and undervalued.
2. Coloniality of Knowledge: Santos’ ‘abyssal lines’ and Quijano’s ‘coloniality’ highlights how dominant systems (from the Global North) render alternative ones invisible and prioritize data that aligns with their interests and conceptualization, further marginalizing infra-data.
3. Standardization Push: The global drive, especially under SDGs, for uniform metrics inadvertently sidelines data that challenges these frameworks. Infra-data’s depth often contradicts these standardized metrics.
4. Algorithmic Biases: As edtech leans on data-driven algorithms, these often reinforce biases by relying on dominant data sets, sidelining infra-data.

Tapping into Infra-Data for Transformative Change– How Do We Proceed?

The exploration of infra-data presents a compelling case for its pivotal role in reshaping global education. This rich mosaic of evidence from the Global South holds the potential to redefine our understanding of education, bridging the gap between grassroots realities and global metrics. As we approach the SDG deadlines, it’s imperative to leverage infra-data for well-crafted educational policy and solutions.

1. Build Pathways between infra-data and official education data regimes, ensuring a seamless flow of information that respects local nuances while aligning with global standards.
2. Re-examine Standards while Building Capacities: Challenge and refine our conceptualizations of what constitutes educational data, ensuring that they are inclusive of the diverse insights infra-data offers. Invest in local initiatives to enhance the quality and scope of data collection, ensuring that infra-data is both robust and representative.
3. Ethical Standards for Data Use: Developing standards and practices includes ensuring the privacy and security of individuals, especially when dealing with sensitive information. Ethical considerations also extend to how the data is interpreted and used, ensuring that it’s not manipulated or taken out of context to serve ulterior motives.
4. Open Sourcing and Stakeholder Collaboration: By making infra-data accessible to researchers, policymakers, and educators globally, we enable learnings from successful interventions to be shared and replicated across different contexts.
5. Technological Advancements: The rapid advancements in technology, especially in data collection, storage, and analysis can make the process of gathering infra-data more efficient, accurate, and widespread.

Given the ever-shrinking time for action, infra-data emerges as more than just an alternative; it’s an imperative. By actively integrating it into our broader educational discourse and paradigms, we can ensure a future that can provide universal and meaningful education to learners across the Global South.

About the Authors:

Hamza Sarfraz is an education data scientist and researcher from Pakistan. He currently manages impact at a learner centric EdTech company. Previously, he was involved with ASER Pakistan and co-authored the country profiles on Pakistan for the GEM Reports 2020 and 2021.

hamzasarfraz1@hotmail.com

Zain Ul Abidin, an education policy researcher and practitioner currently working as a Data Analyst at the University of Glasgow, is interested in global education policy, international development, girls’ education, and Equality, Diversity, and Inclusion. His work on the second chance education program with Idara-e-Taleem-o-Aagahi was recently awarded as one of the top 100 education innovations by HundrED.org.

Zainfarooq138@gmail.com