When we wrapped up our overview of major measurement frameworks in Article 2, one thing was clear: not every framework fits every question. Each offers a different lens for proving value. Yet one model continues to anchor almost every conversation about learning impact: Kirkpatrick’s Four Levels.
It’s the model most learning professionals encounter first and the one many organisations still rely on today. Despite being over 60 years old, it remains foundational because it speaks a language everyone understands: Did people like it? Did they learn something? Did they use it? Did it make a difference?
But while nearly every L&D leader knows the model, far fewer apply it fully or strategically. This article takes a closer look at what the Four Levels are, how to use them effectively, when not to rely on them alone, and what the modern “New World Kirkpatrick Model” adds to the mix.
A Short History: Why Kirkpatrick Still Matters
Donald Kirkpatrick developed his model in the 1950s while completing his PhD at the University of Wisconsin. His goal was simple: to help managers evaluate the effectiveness of training programs in a systematic way. He broke this into four layers: Reaction, Learning, Behaviour, and Results, each answering a different question about impact.
For decades, the model became the backbone of training evaluation worldwide. It was simple enough to use without a statistics background, yet structured enough to produce meaningful insights. Over time, corporate learning matured, but the essence of Kirkpatrick’s idea endured: that training should be judged not by attendance or enjoyment, but by what changes because of it.
Today, in an era of digital platforms, data analytics, and performance-driven learning, Kirkpatrick remains relevant precisely because of its clarity. It provides a shared language that links instructional design, learner experience, and business outcomes, a bridge between learning and performance.
The Four Levels Explained
Each of the Four Levels builds on the previous one. Think of them not as steps to tick off, but as perspectives that connect what happens during learning to what happens afterwards in the business.
Level 1: Reaction — How learners feel
This level measures the learners’ immediate response to the experience. Did they find the content engaging, relevant, and useful? Typical tools include post-session surveys or pulse polls within an LMS.
While often dismissed as “smile sheets,” Reaction data can reveal important signals. If participants felt the content was irrelevant or poorly delivered, engagement and retention will almost certainly drop. The trick is to ask meaningful questions, not just “Did you enjoy it?” but “Was this relevant to your work?” or “Do you feel confident applying what you learned?”
Reaction tells you whether the learning experience resonated. It sets the foundation for everything that follows.
Level 2: Learning — What knowledge or skills were gained
Here, the focus shifts from perception to capability. What did learners actually learn? Assessment can take many forms from knowledge checks, scenario-based exercises, simulations, or peer feedback.
Strong Level 2 measurement captures not just what people know immediately after training, but what they retain and can recall later. Increasingly, modern L&D teams include measures of confidence and commitment, two critical factors that determine whether learning will transfer to work. (We’ll revisit this under the New World Model.)
Level 2 gives you the first tangible evidence that learning has occurred and it offers a predictive signal of whether that learning might translate into performance.
Level 3: Behaviour — What changes in the workplace
This level examines whether learners apply new skills or knowledge on the job. Measuring behaviour is difficult but essential. It often involves observation, 360° feedback, performance data, or digital traces (for instance, system usage or task completion rates).
Here’s why it matters: training only creates value when it changes what people do. A technically perfect course means little if managers still behave the same way after attending.
The best practice is to measure behaviour 30–90 days after the learning experience, when the new habits have had a chance to take root. Pair quantitative indicators (like productivity metrics) with qualitative insights from managers or peers.
Level 3 reveals the bridge between learning and real-world performance.
Level 4: Results — What difference it made
Finally, Level 4 measures the organisational outcomes influenced by the training. Did it improve sales, reduce errors, increase customer satisfaction, or shorten ramp-up time?
This level is often the hardest to measure because many factors affect business results. The key is to define success metrics early, ideally during the design phase, and agree with business stakeholders on what will count as evidence of impact.
By the time you reach Level 4, the story becomes tangible: learning translated into better performance, which created measurable business value.
When and Why to Use the Kirkpatrick Model
Kirkpatrick is best used when you want to evaluate structured learning programs that aim to produce specific performance outcomes for example, leadership development, onboarding, compliance training, or capability uplift.
It’s also ideal when you need to demonstrate alignment between learning and organisational strategy. Each level gives you a piece of the puzzle: engagement (Level 1), capability (Level 2), behaviour change (Level 3), and impact (Level 4). Together, they form a coherent picture of effectiveness.
For executive stakeholders, Kirkpatrick offers a clear narrative: This is how we know training worked and here’s the data from the learner’s reaction to the business outcome.
When Not to Use Kirkpatrick
Not every initiative requires a full four-level evaluation. For small-scale pilots, quick prototypes, or microlearning experiments, the model can feel heavy-handed.
If your goal is rapid iteration or exploratory testing, frameworks like the Success Case Method (SCM) or OKRs may be more practical. Similarly, for data-rich environments where learning is continuous and cross-platform, xAPI and analytics can yield faster, more granular insights.
The main risk is applying Kirkpatrick superficially by collecting Level 1 surveys and calling it “evaluation.” The model only works when you commit to measuring behaviour and results. Without that, it becomes a compliance exercise rather than a performance tool.
Strategic Reasons Organisations Still Use Kirkpatrick
Despite newer models, many organisations return to Kirkpatrick because it is simple, scalable, and credible. It creates a shared vocabulary that unites L&D, HR, and business leaders.
Executives understand it quickly. It mirrors how they already think about performance: satisfaction, learning, behaviour, and results. It’s also flexible enough to integrate with other systems such as ROI calculations, OKRs, or analytics dashboards.
Most importantly, it reinforces a cultural message: learning is not an end in itself. It exists to drive performance and results. That alignment is exactly why the model continues to hold strategic weight, even in data-driven organisations.
Implementation Principles and Advice
Getting Kirkpatrick to work in your organisation doesn’t have to be complicated. Start small, focus on what’s feasible, and build credibility over time. Here’s a structured approach:
- Align with business objectives
Before designing evaluations, clarify the business outcomes your program should influence. Are you improving sales performance, enhancing customer satisfaction, or increasing operational efficiency? Clear objectives ensure your measurement captures meaningful impact. - Start with Level 1 and Level 2
Capture learner reaction and knowledge gain first. Use short surveys, quizzes, or practical exercises. Early wins here are easy to implement and provide immediate insight into program quality and engagement. - Incorporate Level 3 behaviour checks
Observe or track whether learners apply their skills on the job. Partner with managers to gather qualitative or quantitative evidence, such as performance logs, peer feedback, or direct observation. Keep it simple initially focus on key behaviours tied to business objectives. - Link to Level 4 results
Identify one or two business metrics you can influence and connect them to your learning program. Examples include revenue per employee, process efficiency, error reduction, or customer satisfaction. This linkage is what moves L&D from reporting activity to proving impact. - Use incremental pilots
Implement evaluation in a single program before scaling. Pilots allow you to test templates, gather feedback, and refine processes without overwhelming the team or stakeholders. - Communicate findings regularly
Share insights with program designers, managers, and executives. Use concise dashboards or storytelling to highlight both successes and areas for improvement. Demonstrating early wins builds trust and momentum for broader adoption. - Iterate and evolve
Treat measurement as a continuous improvement cycle. Adjust assessments, tools, and focus areas as you learn what captures meaningful impact. Over time, your L&D measurement will grow from simple data collection into a strategic lever that drives performance.
When implemented well, Kirkpatrick becomes more than an evaluation tool, it becomes a continuous improvement system that strengthens both learning design and performance culture.
The New World Kirkpatrick Model
In recent years, Jim and Wendy Kirkpatrick have refined the original model into what they call the “New World Kirkpatrick Model.” The key shift is mindset: start with desired results (Level 4) and work backward.
Rather than treating the four levels as steps to climb, the New World Model treats them as connected elements in a single chain of evidence. It also introduces two critical enhancements: learner confidence and commitment as part of Level 2.
New World Level 2: The Role of Confidence and Commitment
Traditional evaluation often stops at testing knowledge. But research shows that knowing something doesn’t guarantee doing it. Confidence (the belief that you can apply new skills) and commitment (the intention to act) strongly predict whether learning transfers to the workplace.
To measure these, include simple questions such as:
“How confident are you that you can apply this skill in your work?”
“How committed are you to doing so in the next 30 days?”
These can be captured in post-course surveys or follow-up pulses. Combined with manager feedback and behavioural observations, they provide early warning signals about whether learning is likely to stick.
The New World Model also emphasises accountability by ensuring that managers, leaders, and systems support learners in applying what they’ve learned. In other words, transfer isn’t just a learner responsibility; it’s an organisational one.
Templates and Quick Wins
Practical tools can make Kirkpatrick’s model far more approachable and actionable for L&D teams. Simple templates for each level of evaluation save time, ensure consistency, and help teams focus on the metrics that matter. For example, a Level 1 feedback form can be designed not just to capture satisfaction, but also to highlight learner confidence and commitment, setting the stage for Level 2 measurement. Level 3 observation checklists can guide managers in assessing behaviour change without creating administrative burden, while Level 4 scorecards link program outcomes to tangible business metrics.
Quick wins can be surprisingly effective. Start by integrating a single Level 1 survey or Level 2 assessment into one active program. Even small pilots give insight into how data flows, where gaps exist, and how it can influence program design. Sharing these early results with stakeholders builds credibility and demonstrates the immediate value of structured evaluation. Over time, these templates can evolve into a full measurement toolkit that makes Kirkpatrick not just a theoretical framework, but a practical engine for proving impact.
Common Questions and Misconceptions
How soon should we measure behaviour? Usually 30–90 days after training, depending on how quickly skills are applied.
Can Kirkpatrick work for microlearning? Yes, but focus on Levels 2 and 3. Measure small behavioural shifts, not just completion.
Do we need control groups? Not always. Triangulation and trend comparisons are often sufficient in business contexts.
Is Kirkpatrick outdated? No, it’s evolving. The New World Model aligns perfectly with modern, performance-driven learning.
Closing Thought
Kirkpatrick’s model has lasted for a reason. It gives structure to the question every L&D team must answer: Is what we’re doing making a difference?
Used well, it turns learning into evidence of performance improvement. Used poorly, it becomes a box-ticking exercise. The difference lies in intent, whether you’re measuring to prove value or to improve outcomes.
Start simple: map your next program against the four levels, define what success looks like, and gather evidence that tells a meaningful story.
Because in the end, learning only matters if it leads to change, and Kirkpatrick, when applied with purpose, helps you show exactly that.
Running a Pilot with Kirkpatrick in mind? Unlock your practical Pilot Checklists
Take the guesswork out of learning measurement. Download our ready-to-use 12-week and 12-month pilot checklists, built around Kirkpatrick’s model and adapted for South African workplaces. Quickly assess engagement, learning, and behaviour change, and see how outcomes link to real business impact. Start your pilot with confidence today.
Further Reading
- Kirkpatrick Partners — The Kirkpatrick Model: The Standard for Leveraging and Validating Talent Investments
https://www.kirkpatrickpartners.com/the-kirkpatrick-model/
Kirkpatrick Partners, LLC. - Kirkpatrick Partners — An Introduction to The New World Kirkpatrick® Model (Jim & Wendy Kirkpatrick)
https://www.kirkpatrickpartners.com/wp-content/uploads/2021/11/Introduction-to-The-New-World-Kirkpatrick%C2%AE-Model.pdf
Kirkpatrick Partners, LLC. - Kirkpatrick Partners — New World Level 2: The Importance of Learner Confidence and Commitment (Jim & Wendy Kirkpatrick)
https://www.kirkpatrickpartners.com/wp-content/uploads/2024/04/the-new-world-level-2-kirkpatrick-partners-2016.pdf
Kirkpatrick Partners, LLC. - Ardent Learning — What is the Kirkpatrick Model? Learn the 4 Levels of Evaluation
https://www.ardentlearning.com/blog/what-is-the-kirkpatrick-model
Ardent Learning - MindTools — Kirkpatrick’s Four-Level Training Evaluation Model
https://www.mindtools.com/ak1yhhs/kirkpatricks-four-level-training-evaluation-model
MindTools - BusinessBalls — Kirkpatrick Evaluation Method
https://www.businessballs.com/training-assessment-and-quality-assurance/kirkpatrick-evaluation-method/
businessballs.com - EducationalTechnology.net — Kirkpatrick Model: Four Levels of Learning Evaluation (Serhat Kurt)
https://educationaltechnology.net/kirkpatrick-model-four-levels-learning-evaluation/
Educational Technology - Cahapay, M. B. — Kirkpatrick Model: Its Limitations as Used in Higher Education (2021)
https://files.eric.ed.gov/fulltext/EJ1291600.pdf