top of page

FLINT HILLS RESEARCH & EVALUATION

Improvement Science: What the Evidence Says and Why it Matters for Your Organization

  • Writer: Katie Allen
    Katie Allen
  • Jan 20
  • 4 min read

Across healthcare, education, government, and the private sector, leaders are under pressure to deliver better outcomes with finite resources. Over the last three decades, improvement science has emerged as a structured, evidence-informed approach to tackling complex system problems.


Rather than relying on one-off initiatives or heroic leadership, improvement science focuses on how change actually happens in real-world settings. Improvement science offers a robust, evidence-informed approach to tackling complex organizational challenges.


For organizations serious about better outcomes, investing in improvement capability—supported by expert consultancy—represents one of the most credible routes to lasting change. This blog reviews the key strands of the improvement science literature and translates them into practical insights for organizations considering improvement-focused consultancy support.


What Is Improvement Science?

Improvement science is an applied, interdisciplinary field concerned with systematically improving the quality, safety, efficiency, and equity of services. It draws on systems thinking, organizational learning, statistics, behavioral science, and implementation research.


Unlike traditional research, which priorities generalizable knowledge under controlled conditions, improvement science emphasizes:

  • Context-sensitive change

  • Rapid, iterative testing

  • Learning while doing

  • Practical impact over theoretical purity


Prominent definitions (e.g. Batalden & Davidoff, 2007) describe improvement as the combined and unceasing efforts of professionals, service users, and organizations to make changes that lead to better outcomes, better systems, and better professional development.


Historical Foundations

Improvement science has its roots in industrial quality movements, particularly the work of W. Edwards Deming on systems thinking and variation, Joseph Juran on quality planning and control, and Walter Shewhart on control charts and iterative learning. These ideas entered healthcare and public services more fully in the 1990s, notably through the Institute for Healthcare Improvement (Berwick, 1996). Healthcare then became a major catalyst for improvement science due to its complexity, risk, and rich data environment, with landmark reports such as To Err Is Human (Institute of Medicine, 2000) reframing poor performance as a system problem rather than individual failure.


Improvement science is a relatively recent development in education, gaining prominence over the past decade as schools and systems have begun applying disciplined, evidence-informed methods—adapted from healthcare and industry—to address complex, persistent improvement challenges at scale.


Core Concepts in the Literature

1. Systems Thinking

A consistent finding across the literature is that poor performance is usually a property of the system, not the people within it. Improvement science encourages organizations to:

  • Map processes end-to-end

  • Understand interdependencies and feedback loops

  • Address root causes rather than symptoms


This contrasts with compliance-driven or target-led approaches that often generate unintended consequences.


2. Variation and Measurement for Learning

Improvement science places heavy emphasis on understanding variation:

  • Common cause variation (inherent to the system)

  • Special cause variation (due to specific changes or events)


Tools such as run charts and statistical process control (SPC) charts are widely supported in the literature as superior to point-in-time metrics for guiding change (Perla et al., 2011). Crucially, measurement is framed as learning, not performance management.


3. The Model for Improvement and PDSA Cycles

The Model for Improvement, popularized by Langley et al. (2009), remains one of the most empirically grounded frameworks. It asks three questions:

  1. What are we trying to accomplish?

  2. How will we know a change is an improvement?

  3. What changes can we make that will result in improvement?


These are operationalized through Plan–Do–Study–Act (PDSA) cycles. The literature is clear that discipline matters: superficial or poorly designed PDSAs produce little learning.


4. Human Factors and Co-production

More recent work integrates human factors, ergonomics, and co-production (Dixon-Woods et al., 2014; Vincent & Amalberti, 2016). Evidence suggests that sustainable improvement is far more likely when:

  • Frontline staff are active designers, not passive recipients

  • Service users contribute experiential knowledge

  • Psychological safety supports experimentation


5. Context and Adaptation

A major theme in the literature is that interventions do not work independently of context. What succeeds in one organization may fail in another due to differences in:

  • Culture and leadership

  • Capability and capacity

  • Incentives and constraints


This has led to closer links between improvement science and implementation science, particularly around adaptation rather than blind replication (Nilsen, 2015).


What This Means for Organizations

For organizations, the evidence suggests that improvement science is most effective when it is:

  • Strategically aligned to organizational priorities

  • Embedded into leadership and governance

  • Supported by skilled facilitation and coaching

  • Focused on building long-term improvement capability


This is not about importing a methodology, but about changing how the organization learns and adapts.


Let's Talk!

Done well, improvement consultancy is not about bringing in an external expert to tell people what to do. It is about accelerating progress while strengthening the organization from the inside. As an Improvement Science consultant, I:


  • Work as a delivery partner, side by side with leaders and frontline teams, rather than acting as an auditor or inspector

  • Design approaches around your context, priorities, and constraints — not a generic methodology

  • Build internal improvement capability, developing confident leaders and skilled improvers who can sustain change long after the engagement ends

  • Balance pace with rigor, helping you move quickly without cutting corners that undermine learning or results


Contact Flint Hills Research and Evaluation to discuss your organizational improvement needs and to develop a plan tailored to engage key stakeholders and produce lasting positive change.

Key References (Selected)


 
 
 

Comments

bottom of page