#victoriamarsick — Public Fediverse posts

Learning culture: the missing link in global health between learning and performance

Learning culture is a critical concept missing from health systems research.

It provides a practical and actionable framework to operationalize the notion of ‘learning health systems’ and drive transformative change.

Read this first: What is double-loop learning in global health?

Watkins and Marsick describe learning culture as the capacity for change. They identify seven key action imperatives or “essential building blocks” that strengthen it: continuous learning opportunities, inquiry and dialogue, collaboration and team learning, systems to capture and share learning, people empowerment, connection to the environment, and strategic leadership for learning (Watkins & O’Neil, 2013).

Crucially, the instrument developed by Watkins and Marsick assesses learning culture by examining perceptions of norms and practices, not just individual behaviors (Watkins & O’Neil, 2013).

This aligns with Seye Abimbola’s assertion that learning in health systems should be “people-centred” and occurs at multiple interconnected levels.

Furthermore, this research demonstrates that certain dimensions of learning culture, like strategic leadership and systems to capture and share knowledge, are key mediators and drivers of performance outcomes (Yang et al., 2004).

This provides compelling evidence that investments in learning can yield tangible improvements in health delivery and population health.

Learn more: Jones, I., Watkins, K. E., Sadki, R., Brooks, A., Gasse, F., Yagnik, A., Mbuh, C., Zha, M., Steed, I., Sequeira, J., Churchill, S., & Kovanovic, V. (2022). IA2030 Case Study 7. Motivation, learning culture and programme performance (1.0). The Geneva Learning Foundation. https://doi.org/10.5281/zenodo.7004304

As Watkins and Marsick (1996) argue, to develop a strong learning culture, we need to “embed a learning infrastructure”, “cultivate a learning habit in people and the culture”, and “regularly audit the knowledge capital” in our organization or across a network of partners.

While investments in learning can be a challenging sell in resource-constrained global health settings, this evidence establishes that learning culture is in fact an indispensable driver of health system effectiveness, not just a “nice to have” attribute.

Subsequent studies have also linked learning culture to key performance indicators like care quality, patient satisfaction, and innovation.

Why lack of continuous learning is the Achilles heel of immunization

To advance learning health systems, it is important to translate this research in terms that resonate with the worldview of global health practitioners like epidemiologists and to produce further empirical studies that speak to their evidentiary standards.

Ultimately, this will require expanding mental models about what constitutes legitimate and actionable knowledge for health improvement.

The learning culture framework offers an evidence-based approach to guide this transformation.

References

Abimbola, S. The uses of knowledge in global health. BMJ Glob Health 6, e005802 (2021).

Watkins, K. E. & O’Neil, J. The Dimensions of the Learning Organization Questionnaire (the DLOQ): A Nontechnical Manual. Advances in Developing Human Resources 15, 133–147 (2013).

Watkins, K., & Marsick, V. (1996). (Eds.). In action: Creating the learning organization (Vol. 1). Alexandria, VA: ASTD Press.

Yang, B., Watkins, K. E. & Marsick, V. J. The construct of the learning organization: Dimensions, measurement, and validation. Human Resource Development Quarterly 15, 31–55 (2004).

#ChrisArgyris #doubleLoopLearning #globalHealth #healthSystemsResearch #KarenEWatkins #learningCulture #learningHealthSystems #VictoriaMarsick

What is double-loop learning in global health?

Argyris (1976) defines double-loop learning as occurring “when errors are corrected by changing the governing values and then the actions.” He contrasts this with single-loop learning, where “errors are corrected without altering the underlying governing values.”

• Double-loop learning involves questioning “not only the objective facts but also the reasons and motives behind those facts”.
• It requires becoming aware of one’s own “theories-in-use” – the often tacit beliefs and norms that shape behavior – and subjecting them to critical reflection and change. 

This is challenging because it can threaten one’s sense of competence and self-image.

Checking for double-loop learning: ‘Are we doing things right?’ vs. ‘Are we doing the right things?’

In global health, double-loop learning means not just asking “Are we doing things right?” but also “Are we doing the right things?” It means being willing to challenge long-held assumptions about what works, for whom, and under what conditions.

Epistemological assumptions (“we already know the best way”), methodological orthodoxies (“this is not how we do things”), and apolitical stance (“I do health, not politics”) of epidemiology can predispose practitioners to be dismissive of a concept like double-loop learning. 

Learn more: Five examples of double-loop learning in global health

Seye Abimbola is part of a growing community of researchers who argue that double-loop learning is critical for advancing equity and self-reliance in global health systems, because global health tends to overlook its own assumptions.

Is it reasonable to posit that some global health interventions have been driven by unchecked assumptions – assumptions about what communities need, what they value, and what will work in their context? How often have we relied on a one-size-fits-all approach, implementing ‘best practices’ from afar without fully understanding local realities? How do we know to what extent programs have thereby failed to meet their goals, wasted precious resources, and may have even caused unintended harm?

As Abimbola (2021) notes, “double-loop learning goes further to question and influence frameworks, models and assumptions around problems and their solutions, and can drive deeper shifts in objectives and policies.”

For example, affected communities hold vital expertise to mitigate health risks.

However, fully leveraging this potential requires global health professionals to fundamentally rethink their roles and assumptions.

• For research to serve the needs of affected communities, it is likely to be useful to reframe these roles and assumptions to see themselves as “subsidiary” partners in service of “primary” community actors (Abimbola, 2021).
• Institutionalizing double-loop learning requires enabling critical reflection and co-production between health workers, managers and citizens (Sheikh & Abimbola, 2021).
• It also depends on developing the learning capacities of communities and health workers in areas like participatory governance, team-based learning and innovation management.

The next logical question is ‘how’ to implement double-loop learning.

Learning culture is a critical concept missing from health systems research.

It provides a practical and actionable framework to operationalize the double-loop learning notion of ‘learning health systems’ and drive transformative change.

Learn more: Learning culture: the missing link in global health between learning and performance

Further reading

Learning-based complex work: how to reframe learning and development

What learning science underpins peer learning for Global Health?

How do we reframe health performance management within complex adaptive systems?

References

Abimbola, S. The uses of knowledge in global health. BMJ Glob Health 6, e005802 (2021). https://doi.org/10.1136/bmjgh-2021-005802

Argyris, C. Single-loop and double-loop models in research on decision making. Administrative science quarterly 363–375 (1976). https://doi.org/10.2307/2391848

Argyris, C. Double-loop learning, teaching, and research. Academy of Management Learning & Education 1, 206–218 (2002). https://www.jstor.org/stable/40214154

Kabir Sheikh & Seye Abimbola. Learning Health Systems: Pathways to Progress. (Alliance for Health Policy and Systems Research, 2021).

Image: The Geneva Learning Foundation Collection © 2024

Share this:

• WhatsApp
• Email
• LinkedIn
• Facebook
• Twitter
• Telegram
• Reddit
• Print
• Pocket
• Mastodon
• Nextdoor

#ChrisArgyris #doubleLoopLearning #globalHealth #healthSystemsResearch #KabirSheikh #KarenEWatkins #learningCulture #learningHealthSystems #performance #SeyeAbimbola #VictoriaMarsick

Ideas Engine: What is The Geneva Learning Foundation’s insights mechanism?

It’s a cliché to claim that data is the “new oil”, a resource to be mined. We collect it from the field, refine it with experts, and utilize it for decision-making. However, we rarely ask what this extractive model does to the workers and communities that provide the raw materials. This is a summary of how and why we developed the Ideas Engine to collect and share insights.

The flow of data remains largely one-way. We ask local actors to report on vaccination coverage, disease outbreaks, or supply shortages. Yet, all too often, this valuable information travels up the chain without ever returning to the people who generated it in a way they can use.

What if the act of reporting was, in itself, an act of learning? What if the input mechanism was designed not just to feed a database, but to inform the practitioner? What if this recognized the significance of qualitative experiences that are usually dismissed as anecdotes? 

This shift in perspective is the driving force behind The Geneva Learning Foundation’s Ideas Engine, first launched in July 2020 with a group of more than 600 practitioners who designed the COVID-19 Peer Hub with support from the Bill & Melinda Gates Foundation (BMGF).

This mechanism is helping us move beyond the traditional survey model to create a system of reciprocal value. Every piece of data shared becomes a tool for empowerment, connection, and locally-led change.

Ideas Engine: moving beyond mining the frontline

Epidemiologists are trained to dismiss experience as anecdotal, to minimize bias, and to extract clean data. We treat the local actor as a sensor or a passive instrument to measure coverage or disease incidence. But a local actor is not a sensor. She is a professional with the capacity to think, act, and learn. And yet, data reported by local actors are treated with suspicion, generally assumed to be unreliable for multiple reasons.

When we treat a community volunteer or a district medical officer merely as a source of data, we do more than miss the context. We strip them of their agency. We reduce a thinking, adapting professional operating in a complex adaptive system to an anonymous row in a dataset.

This is an epistemic injustice. It assumes that knowledge resides in the center, with experts who analyze the data, while the periphery become an anonymous source or informant.

When we treat people and communities as data sources, we also fail to capture the tacit knowledge that explains the numbers. We miss the story of how a nurse in Kano negotiated with a community leader to allow vaccinators entry. We miss how a district officer in Bihar adapted cold chain logistics during a flood.

The Insights mechanism that led to developing the “Ideas Engine” is not a survey tool designed to extract information for the center. It is a pedagogical pattern designed to build power at the periphery. It supports local actors’ inherent capacity to learn from each other, while offering global actors a rare opportunity: the chance to listen, to act on what they hear, and to question governing assumptions that drive global strategies.

Our Insights mechanism is designed to capture this layer of reality. It operationalizes what learning theorists like Diana Laurillard describe as a conversational framework, but applies it outside classrooms and at a massive scale. Instead of a teacher-student dialogue, we facilitate a peer-to-peer dialogue across borders. This draws on George Siemens’s connectivism, where learning happens by connecting nodes of information across a network. We then add a critical layer of structure to ensure those connections lead to action. This embodies Cope and Kalantzis’s vision of active knowledge production, where the learner is not a consumer of content, but a creator of it. Last but not least, we draw on the insights from the work of Karen E. Watkins and Victoria Marsick to map the capacity for change or “learning culture” that set outer boundaries that local actors operate within.

This mechanism remixes these theoretical frameworks to life on the outer cusp of chaos. It operates in humanitarian emergencies, disasters, war zones, and extreme poverty, engaging tens of thousands of participants where traditional systems fracture. 

Reciprocity as justice, not transaction

In traditional marketing, there is a concept called give-to-get. You give a free resource to get an email address. This is transactional. Our philosophy is different. We believe that giving back is a requirement of justice.

When a health worker in a conflict zone takes thirty minutes to share a story about overcoming vaccine hesitancy, they are performing unpaid labor for the global good. If we do not return that value to them rapidly and in a usable form, we are participating in the same extraction we claim to oppose.

Learn more: Why answer Teach to Reach Questions?

Our Insights mechanism is therefore built on a specific architecture of reciprocity. It cycles value back to the contributor at every stage of the process. This ensures that the mechanism serves the practitioner first, and the hierarchy is positioned in support of the practitioner. This distinct ethical framework is what allows us to maintain high levels of engagement and trust over time.

The architecture of the Ideas Engine: from reflection to action

The mechanism is a complex assembly of pedagogical scripts, technical workflows, and community engagement loops. It functions as the central operating system for our learning programs, feeding both the Teach to Reach events and the Impact Accelerator.

1. The input: reflections, not reporting

Standard data collection asks for statistics. How many children did you vaccinate? This triggers compliance. Our questions ask for narratives. Tell us about a time you faced a challenge. What did you do?

This phrasing is intentional. It forces the user to pause and reflect on their own practice. This is metacognition. It transforms them from a data subject into a knowledge producer.

2. The immediate return: collections of experiences

Our insights team reads every contribution. The team then does the grueling work of producing a collection of Shared Experiences. This is a compendium with hundreds and sometimes thousands of peer stories. It is filtered only to remove nonsensical or AI-generated content.

We strive to share this back with the community as quickly as possible. This validates tacit knowledge. It tells the health worker that their experience matters enough to be shared with the world rapidly. It is also that a health worker facing a cholera outbreak today is more likely to benefit if the latest experiences are shared when and where they are needed, not on a scholarly publishing calendar that may take months or years. (Our process includes peer feedback, and we posit it actually resolves some of the challenges being faced by academic publishing.)

3. The synthesis: thematic insights reports

While the raw collection is fast, we then use more conventional qualitative research techniques to produce thematic insights reports, also known as “eyewitness reports”. Each report distills dozens, hundreds, or thousands of contributions into short summaries of what we learned from them, on a specific topic or challenge. Written for the community, they identify patterns that no single individual could see on their own. These reports also turn out to be surprisingly relevant and useful for non-local actors.

4. The dialogue: dynamic event-driven knowledge translation

Knowledge in action is dynamic, by definition. The Ideas Engine is about turning knowledge into action. This is why we host Insights Live. These are rapid-fire livestreamed sessions where the data comes alive, with contributors, guides on the side, and anyone else interested joining to discuss how they are putting to use what we are learning together.

We invite the contributors themselves to take the floor as our guests of honor. A lot of what happens in these live session – who speaks, what we learn – we cannot and do not predict in advance. It is emergent. This is more akin to jazz improvisation, rather than the rigid classical music orchestration of presentation webinars. We invite global partners and funders to listen. This reverses the usual power dynamic. We then turn these livestreamed events into podcasts. This ensures that even those with low bandwidth or no time to watch a screen can access the learning.

5. The application: closing the loop

Knowledge is useless if it cannot be shared. This is why we provide tools for dissemination. For example, we prepare short slide decks that contributors can use to present insights to their colleagues and teams.

Crucially, this includes a feedback facility. We track not just who downloaded the deck, but who presented it and what their colleagues said. This allows us to measure the ripple effect of the insight, including actual use and, in some cases, how the use of an insight led to changes in practice and tangible improvements in outcomes.

Does the Ideas Engine actually make a difference?

Does this actually work? Is it better than a survey? The data suggests yes.

In an independent analysis by the University of South Australia’s Centre for Change and Complexity in Learning, researchers examined our Ideas Engine. This was a core component of this mechanism during the COVID-19 Peer Hub. The report revealed the scale of engagement that this proprietary method generates.

• Scholars contributed 1,103 ideas and 3,061 comments. This is an average of 2.77 comments per idea.
• 80.2% of participants reported using the Ideas Engine.
• Of those who used it, 92.9% reported finding ideas that were useful for their work.
• Perhaps most importantly, the analysis of citations showed that two-thirds of the citations in action plans were to ideas from peers working at different levels of the health system.

This proves that the mechanism does not just collect data. It successfully bridges the gap between knowledge and action by connecting practitioners across hierarchies.

Photo: The Geneva Learning Foundation Collection © 2020

This article was updated on 6 January 2026 to reflect what we have learned since 2020.