Global Learning Stagnates While Some Nations Advance

Since 2017, average learning outcomes across OECD countries have not only stagnated but have also shown a reduction, indicating a puzzling backward trend. This decline suggests that current educational approaches are failing to maintain, let alone improve, foundational learning levels globally. The PISA results consistently reflect this disheartening pattern, prompting questions about the effectiveness of existing educational policies.

These countries are moving in the opposite direction of the OECD average, suggesting successful strategies are being implemented. Their success stems from a strong focus on foundational learning and a serious commitment to effective implementation. This progress indicates that targeted efforts can yield positive results even in challenging global contexts.

This finding is significant because quantity measures are generally easier and more cost-effective to collect at scale. The F-test revealed no statistically significant difference in the predictive value for learning outcomes between the two types of measures, suggesting that complex quality indicators often introduce noise or are difficult to implement consistently.

This involves a strategic redesign of indicators, moving beyond tracking simple access to focusing on instructional implementation. The new approach retroactively designs indicators based on the specific actions they are intended to enable, integrating classroom observers and various stakeholders into the decision-making process.

This redesign shifts the platform's focus from compliance tracking to actively aiding classroom instruction. The strategy involves strengthening existing government systems instead of creating parallel structures, emphasizing specific and observable indicators for frontline supervisors.

Endogenous selection bias significantly influences technology adoption, meaning individuals who choose technology often differ fundamentally from those who do not. Mandating technology adoption can lead to a 'voltage drop' in student achievement, as performance gaps emerge between 'compliers' (those who adopt only when mandated) and 'always takers.' This suggests that simply increasing take-up rates does not necessarily translate into proportional improvements in learning outcomes.

A study highlights the significant heterogeneity among teachers in large school systems regarding technology adoption. While senior leadership often drives technology policy, younger teachers typically possess greater digital fluency. Leveraging this age diversity by empowering younger teachers as 'tech captains' can facilitate intra-school knowledge transfer and serve as a crucial laboratory for scaling educational ideas.

Effective scaling of educational programs requires a delicate balance between broad reach (quantity) and deep implementation fidelity (quality). Quantity is visualized as a growing circle, while quality is represented by the intensity of its shade. The objective is to avoid creating large, 'hollow' programs that lack substantive impact or tiny, undelivered initiatives. Achieving this balance means finding a middle ground that ensures both widespread access and meaningful engagement.

A simple grouping guide, costing $0.76 per child, demonstrated superior cost-effectiveness compared to coaching, which cost $1.87 per child. The grouping guide yielded 15 standard deviations of gain for every $100 spent, while coaching produced 10 standard deviations. Although coaching is less cost-effective in direct gains, it plays a vital role as a reinforcement mechanism, gradually increasing implementation fidelity over time.

In Namibia, the prohibitive costs of physical school visits due to remote geography led to an innovative solution: replacing the second school visit with a coaching call. This change reduced program costs by a substantial 33%, cutting one-third of the total delivery cost for the Tarl program. Crucially, learning outcomes remained consistent between the groups receiving physical visits and those receiving coaching calls, demonstrating the flexibility and efficiency of remote delivery methods.

Research ownership was not confined to a single individual or department; rather, it became a collective responsibility. Teams were diversified to include members with expertise in male, pedagogy, and field operations, and every loop of the research process involved training more individuals to conduct research effectively.

The idea emerged from the need to close implementation gaps, recognizing that Ghana already has cluster area supervisors (CISOs). This approach aims to utilize existing human resources to provide localized and timely support, thereby maximizing efficiency and effectiveness within the educational system.

The primary constraints limiting the frequency of rapid, wide-scale testing are financial resources. Beyond monetary considerations, aligning with government priorities is essential for any testing initiative to proceed. Furthermore, testing must seamlessly fit within the existing school calendar to avoid disruption. These factors collectively restrict the ability to conduct frequent and extensive educational assessments.

Iterative testing allows for early screening and elimination of weak features before significant commitment to interaction testing. Crossover design effectively separates true treatment effects from exposure duration, providing clearer insights into program efficacy. Finally, embedding programs within government structures ensures both rigor and feasibility for sustainable expansion.

The study revealed massive improvements in foundational skills, amounting to 40% of a standard deviation, following intervention. However, these gains remained narrow, showing hardly any impact on literacy paragraph reading. Crucially, after two years, there was little transfer of these foundational gains to other skills, particularly in mathematics where specific targeting occurred. This suggests that while targeted interventions can boost specific foundational abilities, broader cognitive transfer does not automatically follow.

Student attendance significantly hampered program effectiveness, with only half of the tracked students regularly attending school. Furthermore, only two-thirds of the students present were actually exposed to the catch-up program, meaning only one-third of the intended exposure was achieved overall. This 'light exposure' limited the program's full potential, while continuous professional development, despite its costs, did not improve outcomes and effectively doubled program expenses.

Schools identified as bottom-performing experienced a 5% increase in infrastructure funding. These investments encompassed essential facilities such as computer labs, bathrooms, and libraries. This demonstrates that governments utilize transparent information to allocate resources strategically, aiming to improve underperforming educational areas.

Many middle-tier bureaucrats involved in the study possessed limited formal teaching experience and were not pedagogy specialists. This lack of specialized knowledge means that complex instructional support tasks are frequently simplified or reduced to 'codifiable' administrative processes. The prevailing bureaucratic culture often prioritizes compliance and output metrics over actual pedagogical outcomes, further exacerbating the gap in instructional quality.

In Gauteng, the initial focus was successfully achieving credibility in data production. The current challenge lies in transforming this established data credibility into tangible, real-world consequences for schools. Subject advisors need to evolve from passive data collectors to active interpreters of information, driving meaningful change. To achieve this, small-scale, focused interventions are recommended over broad-brush strategies, ensuring targeted and effective action.

In South Africa, middle-tier educational staff are appointed as subject matter specialists, not political appointees. Every individual appointed to a middle-tier position must hold qualifications in either home languages, foundation phase education, or mathematics. Furthermore, securing the head of school's buy-in is deemed vital for the successful implementation and sustained impact of any educational program.

This trend signals a notable strengthening of the public education system.

Integrating pre-primary grades into government primary schools before the first grade has demonstrated positive consequences for learning outcomes and helps retain children within the public system. States like Punjab and Himachal successfully implemented this before national policy. Providing early childhood education within government schools aids parents in keeping their children in public education and ensures age-appropriate entry into the first grade, which is cognitively essential for foundational learning.

Australia's Department of Foreign Affairs and Trade (DFAT) has formally joined the What Works Hub as a key funding partner, signing an agreement with Oxford in June. This commitment extends support for the Hub's global education research work, specifically expanding its reach into the Indo-Pacific region. New target countries for the Hub include Papua New Guinea and the Philippines, aligning with Australia's management of over 130 education and skills programs across the Indo-Pacific.