Data Storytelling in a World of Knowledge Asymmetry
https://www.thetransmitter.org/
Two reports landed the same month. Their contexts could not be more different. The problem they expose is the same.
Unlocking the Power of Education Data: Policy Solutions to Bridge Foundational Learning Gaps in Kenya [Policy Brief] Gachoki, & Mutisya, and Rethinking the Future of Data Science Education: A Case for Thoughtful Design to Integrate AI into the College Classroom, Yarnall et al.
For years, educators and policymakers have been told that the central challenge of education systems is data scarcity — that we simply do not know enough about what students are learning. Increasingly, the evidence suggests the opposite. The problem is not the absence of data. It is the inability of systems to turn evidence into instructional improvement.
More than a decade ago, Taking College Teaching Seriously argued that improving learning outcomes requires more than collecting evidence about students — it requires educators to examine and codify practice and translate that evidence into changes in teaching.
Yarnall and colleagues asked why data storytelling — the skill of structuring meaning from analysis — keeps getting squeezed out of data science courses in US higher education. Gachoki and Mutisya documented why foundational learning data in Kenya, abundant and longitudinal, fails to change what happens in classrooms.
One paper asks how to teach people to build the story. The other asks why the story never reaches the people who need to act on it.
Both are data storytelling problems.
That gap matters. It shapes which problems get centered when AI in education is designed, which evidence informs reform, and which learners’ outcomes are treated as the measure of what works.
The Discipline Behind the Design:
Yarnall and colleagues observed something specific before proposing anything. Faculty spend disproportionate time fielding repetitive syntax questions. Students struggle to structure meaning from analysis. There is not enough time for the iterative, relational feedback that transforms technical output into insight.
Before proposing AI, the researchers mapped where learning actually breaks down. They observed classrooms. They interviewed faculty and students. They identified cognitive bottlenecks.
Only then did they ask what role AI should play in addressing these bottlenecks.
The goal of introducing AI is not to replace teaching. It is to strengthen teaching in ways grounded in evidence. AI is valuable only insofar as it strengthens teaching effectiveness and improves learner outcomes — not because it is novel, fast, or scalable, and not when it is deployed primarily to produce dashboards divorced from the work of learning. If AI absorbs repetitive, low-level load and frees instructors to do the analytic, relational, and interpretive work that improves understanding, it serves pedagogy. If it displaces that work, it weakens outcomes.
This is Purposeful Pedagogy operating at classroom scale.
When the Loop Never Closes:
Kenya has national assessments. It has longitudinal literacy and numeracy data from hundreds of thousands of children. It shows which learners are unable to read a Grade-3-level text and which cannot solve a basic mathematics problem. What it does not have is a reliable connection between that knowledge and improvements in teaching.
Assessment results circulate. Reports are written. Frameworks exist. Yet, teacher preparation has not meaningfully shifted. The Kenya School Readiness Assessment framework remains unimplemented. Platforms are siloed. Data is often inaccessible — even to the ministry itself. Officials often lack the analytic capacity to interpret what they see.
The data goes into reports. The reports go into libraries. The classrooms carry on.
This is not a data deficit. It is a systems failure. Evidence exists but does not reliably modify instruction, program design, or policy. The loop between assessment and action never fully closes.
This is Purposeful Pedagogy failing at systems scale.
Through the Lens of Sustainable Learning:
Two pillars of the Sustainable Learning Framework are directly implicated here: Purposeful Pedagogy and Digital Stewardship.
Purposeful Pedagogy requires that pedagogical practices be based on the best available evidence rather than tradition, personal judgment, or convenience. Evidence is not decorative. It is operational. Educators must compile, analyze, and use objective evidence to inform academic program design, guide modification of instructional techniques, and shape policy decisions. Yarnall's work is the discipline in action. Gachoki and Mutisya reveal what happens when it is absent at scale.
Digital Stewardship adds another layer. Institutions must govern the digital environments they create. Data must be accessible, interoperable, and connected to decision pathways. Without governance, even high-quality evidence remains inert.
And then there is the asymmetry — the unequal distribution of capacity to generate, access, interpret, and act on evidence.
The Kenya brief notes that foundational learning research concentrates in urban areas, privileges literacy over numeracy, and systematically excludes children in arid and semi-arid lands, children with disabilities, children out of school, and refugee children. The data ecosystem does not merely fail to reach these children. It often fails to see them.
This is not a measurement glitch. It is a Democratic Engagement failure. When entire populations are structurally absent from the evidence base, Purposeful Pedagogy becomes impossible. You cannot build on evidence that does not count the learners most in need of instructional response.
The global AI-in-education conversation is being shaped within one knowledge infrastructure. The evidence emerging from more fragile systems circulates within another. That asymmetry determines whose problems get centered — and whose solutions scale.
A Question Worth Holding:
At what point in your system does evidence stall?
In the classroom? In the district? In the ministry? Where does data stop shaping decisions?
We do not suffer from a shortage of data. We suffer from systems that do not know how to build on it. Purposeful Pedagogy requires more than collection and more than analysis. It requires that evidence actually modify instruction — that it inform program design, teacher preparation, and policy decisions in visible, iterative ways. Digital Stewardship requires infrastructure that connects data to those decision pathways, rather than accumulating reports that sit adjacent to practice.
If evidence does not alter teaching and learning — whether in a college classroom or a national ministry — it is not evidence in action. It is reporting.
And reporting, however sophisticated, does not educate children.
Learning systems improve only when evidence becomes part of the design of teaching itself.
The asymmetry that matters most is not between two papers or two funding streams. It is between systems capable of translating evidence into improved learning and systems that cannot. Until that gap is addressed, AI will scale visibility faster than it scales understanding — and dashboards will multiply faster than outcomes improve.
The data already exists. What is missing are systems designed to learn from it.
Resources
On systems and data governance
Research and initiatives examining how education systems organize, govern, and use data to inform policy and practice.
Unlocking Data Initiative (Zizi Afrique Foundation).
On pedagogy and AI design
Work examining how data analysis, interpretation, and storytelling are taught — and how AI may support those processes.
Mellow, G. O., & Woolis, D. (2010). Taking College Teaching Seriously: Pedagogy Matters.
Yarnall, L., et al. (2025). Laying Foundations for Scalable Coaching for Data Storytelling.
On data storytelling research
Scholarship defining how data, narrative, and visualization combine to translate analysis into meaning and decision-relevant insight.
Kosara, R., & Mackinlay, J. (2013). Storytelling: The Next Step for Visualization. IEEE Computer.
Hullman, J., et al. (2023). Telling Stories with Data: A Systematic Review.
Data storytelling in practice
Organizations demonstrating how narrative data analysis can inform public understanding and policy conversations.
Tools and methods for practitioners
Resources for educators, analysts, and policymakers translating datasets into narratives that inform decisions.
Knaflic, C. Storytelling with Data. Instructor Resource Hub.
Datawrapper. Visualization and storytelling platform widely used in global data journalism.
Flourish. Interactive data storytelling and visualization tools.
Hands-On Data Visualization. Open textbook and methods for building narrative visualizations.
Evergreen Data. Evaluation and data storytelling methods for education and nonprofit organizations.
