e-trainia
Back
Education Blog

What Is a Kirkpatrick Model? The 4 Levels of Training Evaluation Explained

Learn what the Kirkpatrick Model is and how its 4 levels — Reaction, Learning, Behavior, and Results — help you evaluate whether corporate training actually works.

M
Milan
10/15/20266 min read
What Is a Kirkpatrick Model? The 4 Levels of Training Evaluation Explained

Introduction

Most training programs get measured by one question: did people finish the course? But completion doesn't tell you whether training actually worked. Did learners enjoy it? Did they retain the material? Did it change how they behave on the job? Did it move any business results at all?

The Kirkpatrick Model was built to answer exactly those questions. Since the 1950s, it's remained one of the most widely used frameworks in corporate learning and development for evaluating training effectiveness — not just whether training happened, but whether it was worth running in the first place.

What Is the Kirkpatrick Model?

The Kirkpatrick Model is a four-level framework for evaluating the effectiveness of training programs, developed by Dr. Donald Kirkpatrick in 1959. Each level measures a different, increasingly meaningful aspect of training impact — starting with how learners felt about the training, and ending with whether it produced measurable business results.

The model is typically used in order, with each level building on the one before it. Most organizations measure Levels 1 and 2 consistently, while Levels 3 and 4 — the ones that actually prove business value — are measured far less often, despite being the most useful.

Level 1: Reaction

What It Measures

Whether learners found the training engaging, relevant, and worthwhile. This is the most immediate and easiest level to measure, typically gathered right after a course ends.

How to Measure It

  • Post-course satisfaction surveys
  • Net Promoter Score (NPS) for the course
  • Star ratings or feedback forms
  • Qualitative comments about content quality or delivery

Why It Matters (and Its Limits)

A positive reaction doesn't guarantee learning happened — plenty of well-liked courses fail to actually build skill. But a consistently poor reaction score is a reliable early signal that something in the content or delivery needs fixing before you even get to measuring deeper impact.

Reaction data tells you whether people liked the training. It says nothing about whether they can actually do the job better because of it.

Level 2: Learning

What It Measures

Whether learners actually gained the knowledge, skills, or confidence the training was designed to build. This moves beyond opinion into measurable comprehension.

How to Measure It

  • Pre- and post-training assessments or quizzes
  • Skills demonstrations or practical tests
  • Certification exam pass rates
  • Comparing scores before and after training to isolate actual knowledge gain

Why It Matters

This is the first level that confirms training content is actually effective at transferring knowledge — not just well-received. A course with high satisfaction scores but low assessment scores usually points to content that's entertaining but not instructionally effective.

Level 3: Behavior

What It Measures

Whether learners are actually applying what they learned back on the job. This is where evaluation gets significantly harder — and significantly more valuable — because it requires observing real work performance, not just test results.

How to Measure It

  • Manager observations and performance reviews conducted weeks or months after training
  • 360-degree feedback from peers or direct reports
  • On-the-job assessments or spot checks
  • Behavioral data tied to the specific skill trained (e.g., call quality scores after a customer service course)

Why It Matters

This is the level most organizations skip — it takes longer to measure and requires cooperation from managers, not just L&D. But it's also the level that separates training that "worked" on paper from training that actually changed how people do their jobs.

Level 4: Results

What It Measures

Whether the training produced a measurable impact on business outcomes — the metrics leadership actually cares about, like productivity, quality, retention, sales, or cost reduction.

How to Measure It

  • Comparing business KPIs before and after training rollout (e.g., error rates, customer satisfaction scores, sales figures)
  • Reduction in safety incidents following compliance or safety training
  • Employee retention rates following onboarding or leadership development programs
  • Cost savings from reduced errors, rework, or support tickets

Why It Matters

This is the level that justifies training budget. It's also the hardest to isolate, since business results are influenced by many factors beyond training alone — which is why it's often measured through correlation and trend comparison rather than a single clean before/after number.

How to Apply the Kirkpatrick Model in Your Organization

  • Start with Level 1 and 2 as your baseline — every course should have a post-course survey and a knowledge check at minimum
  • Pick a few high-priority courses for Level 3 evaluation — don't try to measure behavior change for every course; focus on the training programs with the biggest business stakes (compliance, sales, leadership development)
  • Connect Level 4 to existing business metrics — rather than inventing new measurements, tie training evaluation to KPIs your business already tracks (retention, error rates, sales)
  • Use LMS and performance data together — most LMS platforms handle Levels 1 and 2 natively through built-in surveys and quizzes; Levels 3 and 4 typically require pulling data from HR systems, CRM, or performance management tools
  • Report on the level that matches your audience — L&D teams often care about Level 2, while leadership usually only wants to see Level 4

Common Mistakes When Using the Kirkpatrick Model

  • Stopping at Level 1 — measuring only satisfaction and assuming that means training "worked"
  • Treating each level as separate rather than connected — the levels are meant to build a chain of evidence, not stand alone
  • Trying to measure Level 4 for every course — not every training program needs a full ROI analysis; reserve it for high-stakes programs
  • Ignoring Level 3 because it's harder — behavior change is often the most important evidence that training actually worked, even though it takes longer to observe

The Kirkpatrick Model exists because "training happened" and "training worked" are two very different claims. Reaction and Learning data are useful starting points, but Behavior and Results are what actually prove training moved the needle — and they're often the levels most worth the extra effort to measure, especially for your highest-stakes training programs.

If your current evaluation stops at a satisfaction survey, the Kirkpatrick Model offers a clear next step: pick one important training program, and try measuring all four levels for it. The gap between what Level 1 tells you and what Level 4 reveals is usually where the real insight lives.