The Kirkpatrick Model of Training Evaluation: A Complete Guide

The Kirkpatrick Model is a four-level framework for evaluating the effectiveness of training programs. Created by Donald Kirkpatrick, Ph.D., in 1959, it measures outcomes across four sequential levels: Reaction, Learning, Behavior, and Results. It remains one of the most widely used training evaluation models in the world, applicable to in-person, online, and blended learning formats.

Who Was Donald Kirkpatrick and Why Does His Model Still Matter?

Donald Kirkpatrick was a professor emeritus at the University of Wisconsin who first published his four-level evaluation framework in a series of articles in 1959. He later formalized it in his 1994 book Evaluating Training Programs. After Donald Kirkpatrick's passing, his son James Kirkpatrick and daughter-in-law Wendy Kirkpatrick continued developing the approach through Kirkpatrick Partners. The most recent major update to the model was released in 2016, introducing what Kirkpatrick Partners calls the Kirkpatrick Four Levels with a stronger emphasis on starting evaluation design from Level 4 (Results) and working backward.

The model's endurance comes from its simplicity and its logical hierarchy. Organizations can apply it regardless of industry, training format, or program length. It is particularly relevant for learning and development (L&D) teams that need to demonstrate training return on investment (ROI) to leadership.

What Are the Four Levels of the Kirkpatrick Evaluation Model?

The Kirkpatrick model of training evaluation organizes measurement into four increasingly complex levels. Each level builds on the one before it, creating a hierarchy that moves from immediate participant experience to long-term business impact.

Level 1: Reaction

Level 1 measures how participants respond to a training program immediately after it ends. The goal is to understand whether learners found the content relevant, engaging, and well-delivered. The most common tool at this level is a post-training satisfaction survey or "smile sheet." While reaction data alone cannot confirm that learning occurred, low satisfaction scores are a reliable signal that something needs to change in delivery or design.

Useful Level 1 questions include:

• Was the content relevant to your daily work?
• How would you rate the clarity of the instruction?
• What would you change about this program?

Level 2: Learning

Level 2 assesses the knowledge, skills, attitudes, confidence, and commitment participants gained during training. Measurement tools at this level include pre- and post-assessments, quizzes, practical demonstrations, and simulations. The gap between pre-training and post-training scores indicates how much learning actually took place.

For digital skills training in particular, Level 2 evaluation can include software simulations or scenario-based tests that replicate real tasks the learner must perform.

Level 3: Behavior

Level 3 is where many organizations struggle. It measures whether learners transfer their new knowledge and skills into on-the-job behavior. Because behavioral change does not happen overnight, this evaluation typically takes place weeks or months after training ends. Methods include manager observation, 360-degree feedback, performance reviews, and follow-up surveys.

A critical factor at Level 3 is what Kirkpatrick Partners calls "required drivers" - the reinforcement, encouragement, and reward structures in the workplace that either support or undermine behavioral change. Without those conditions in place, even high-quality training rarely translates into lasting behavior change.

Level 4: Results

Level 4 measures the degree to which the training produced the organizational outcomes it was designed to achieve. These outcomes vary by program goal: reduced error rates, higher sales figures, faster onboarding times, lower support ticket volumes, or improved compliance scores. This is the level most closely associated with training ROI.

Kirkpatrick Partners advise starting the evaluation design at Level 4, before training is even built, so that the program is designed backwards from the desired business result. This approach is often called the "New World Kirkpatrick Model."

How Does the Kirkpatrick Hierarchy Work in Practice?

The Kirkpatrick hierarchy is best understood as a chain of evidence. Positive reactions (Level 1) create the conditions for learning (Level 2). Genuine learning enables behavioral change (Level 3). Sustained behavior change produces measurable results (Level 4). Breaking the chain at any level weakens the case for training effectiveness.

In practice, most organizations measure Level 1 consistently, many measure Level 2, fewer measure Level 3, and only a minority regularly measure Level 4. The difficulty and cost of measurement increases at each level, which is a key limitation of the model discussed later in this article.

Applying Kirkpatrick Evaluation in Digital Learning Environments

The Kirkpatrick model of evaluation maps naturally onto digital and technology training contexts. For organizations rolling out new software, each level has a specific application:

• Level 1: Survey employees on their experience of the software training, whether delivered via e-learning modules, instructor-led sessions, or in-app guidance.
• Level 2: Test whether employees can correctly complete key tasks in the new system using assessments or task simulations.
• Level 3: Monitor actual usage data after go-live to see whether trained behaviors appear in daily work, such as reduced help desk calls or faster transaction completion.
• Level 4: Tie usage metrics to business outcomes, such as process efficiency gains, error reduction, or compliance rates.

Digital adoption platforms generate behavioral data that can directly feed Level 3 and Level 4 evaluation. When employees interact with in-app guidance, anonymized usage logs show whether trained behaviors are being applied consistently across user groups. This kind of data removes much of the guesswork traditionally associated with the upper levels of the Kirkpatrick training evaluation model.

"A platform like Lemon Learning enables digital adoption but also, as a by-product, the real-time observation of behaviours through anonymised logs by profile."

For organizations seeking a structured approach to software training measurement, Lemon Learning's learning and development solution connects in-application guidance with usage analytics that support Kirkpatrick-aligned evaluation at every level.

How to Use the Kirkpatrick Model in Your Training Strategy

Using the Kirkpatrick training evaluation model effectively requires planning before training design begins, not just after it ends.

1. Define Level 4 outcomes first. Identify the specific business results the training must produce and how they will be measured.
2. Design backward through the levels. Determine what behaviors (Level 3) are needed to achieve those results, what learning (Level 2) will produce those behaviors, and what training experience (Level 1) will best support that learning.
3. Assign measurement owners. Each level needs a responsible person or team. Level 3 and 4 data often sits with managers or business analysts rather than L&D.
4. Set evaluation timelines. Collect Level 1 data immediately after training. Schedule Level 3 follow-ups at 30, 60, or 90 days post-training. Measure Level 4 results over the timeframe that matches the business goal.
5. Use data to iterate. The model is a continuous improvement cycle. Evaluation findings should feed back into training redesign.

Pairing the Kirkpatrick model with a broader understanding of learning science strengthens design quality. For example, insights from neuroscience-informed training design can help ensure that Level 2 learning is retained long enough to drive the behavioral change measured at Level 3.

The Kirkpatrick Model from 1959 to Today

Donald Kirkpatrick first described the four-level model in 1959 in a series of articles for the Journal of American Society of Training Directors. The framework was widely adopted throughout the 1960s and 1970s as organizations began investing heavily in corporate training. The 1994 book Evaluating Training Programs gave the model its most complete formal statement and introduced it to a new generation of practitioners.

The 2016 update, developed by James Kirkpatrick and Wendy Kauffman Kirkpatrick, reframed the model to start evaluation at the outcome level rather than the reaction level. It also placed greater emphasis on the role of organizational support systems in enabling behavior change, addressing a long-standing criticism that the original model placed too much responsibility on the training intervention itself.

What Are the Limitations and Disadvantages of the Kirkpatrick Model?

No evaluation framework is without drawbacks. The most commonly cited limitations of the Kirkpatrick model are:

• Attribution difficulty. It is hard to isolate training as the sole cause of business results at Level 4. Other factors, such as management changes, market conditions, or process improvements, may influence the same metrics.
• Cost and time at higher levels. Measuring Level 3 and Level 4 outcomes requires sustained effort and resources. Many organizations lack the infrastructure or budget to do this consistently.
• Post-training orientation. The original model focuses on measuring what happened after training, rather than evaluating the design, preparation, or implementation phases before and during delivery.
• No cost-benefit framework. The Kirkpatrick model does not natively calculate financial ROI. Organizations often combine it with Jack Phillips' ROI Methodology, which adds a fifth level to convert Level 4 results into monetary terms.
• Assumes linear causation. The model assumes that positive reactions lead to learning, which leads to behavior change, which leads to results. This chain does not always hold in complex organizational environments.

Understanding these disadvantages of the Kirkpatrick model helps L&D professionals use it more effectively, combining it with other frameworks rather than relying on it in isolation. For a broader view of structured training design, the complete guide to instructional design provides complementary principles for building programs that perform at every Kirkpatrick level.

How Does the Kirkpatrick Model Compare to Other Evaluation Frameworks?

The Kirkpatrick model of training evaluation is most powerful when used alongside design frameworks like ADDIE (Analysis, Design, Development, Implementation, Evaluation) or learning theories like Kolb's experiential learning cycle, which addresses how adults process and retain new skills.

Key Takeaways

The Kirkpatrick model of evaluation gives learning and development professionals a clear, practical structure for measuring training impact at four levels: Reaction, Learning, Behavior, and Results. Developed by Donald Kirkpatrick in 1959 and updated by Kirkpatrick Partners in 2016, it remains a global standard for training evaluation. Its greatest value comes from designing evaluation backward from desired business outcomes, ensuring that every element of a training program is tied to measurable results. While it has recognized limitations, particularly around attribution and upper-level measurement costs, it provides a shared language and a logical evidence chain that connects individual learning to organizational performance.