The Kirkpatrick Model: 4 Levels, Examples, and Free Template

The Kirkpatrick Model is a four-level framework for evaluating the effectiveness of training programs: Reaction, Learning, Behavior, and Results. Donald Kirkpatrick introduced it in 1959, and it remains the most widely used training evaluation model in corporate learning and instructional design.

This guide covers all four levels with worked examples and sample questions, a planning template, the updated New World Kirkpatrick Model (including the 2026 changes), the model's main criticisms, and how it compares to alternatives like Phillips ROI and LTEM.

What Is the Kirkpatrick Model?

The Kirkpatrick Model is a training evaluation framework that measures the impact of a learning program across four levels: Level 1 (Reaction), Level 2 (Learning), Level 3 (Behavior), and Level 4 (Results). Each level answers a different question about whether the training is working, and each level becomes more difficult (and more valuable) to measure than the one before it.

The model is most often applied to corporate training and eLearning programs, but it's flexible enough to evaluate any kind of learning intervention. Modern practitioners, including the Kirkpatrick family's own organization, treat the model as cyclical rather than linear, and they plan in reverse: define the Level 4 business results first, then work backward to design the training itself.

The Four Levels at a Glance

Here's how the four levels compare across the questions they answer, the methods used to measure them, and when to do that measurement.

Level	Question Answered	Common Methods	When to Measure	Difficulty
1. Reaction	Did learners find the training engaging, relevant, and useful?	Post-training surveys, pulse checks, focus groups	During and immediately after training	Low
2. Learning	Did learners gain the intended knowledge, skills, attitudes, confidence, and commitment?	Quizzes, tests, demonstrations, simulations, pre/post assessments	During and at the end of training	Moderate
3. Behavior	Are learners applying what they learned on the job?	Observation, performance data, 360-degree feedback, supervisor reviews	Starting within a few weeks of training; ongoing	High
4. Results	Is the training producing measurable outcomes for the business?	KPIs, ROE, ROP, Contributive ROI, leading and lagging indicators	3 to 12 months post-training, ongoing	Very High

How Widely Used Is the Kirkpatrick Model?

Kirkpatrick is the dominant evaluation framework in corporate L&D by a wide margin, even though most organizations only apply the lower levels. The numbers tell both stories at once:

Statistic	Figure	Source
Organizations that use Kirkpatrick as their learning evaluation framework, making it the most common by far	71%	ATD Research, 2025 (n=222)
Organizations that evaluate the business results (Level 4) of their learning programs	35%	ATD, Evaluating Learning: Getting to Measurements That Matter, 2016 (n=199)
Practitioners in the Kirkpatrick certified community worldwide	14,000+	Kirkpatrick Partners

That gap between the 71% who use the model and the 35% who actually measure business results is the single most important thing to understand about training evaluation in practice. We'll come back to it in the criticisms section.

Who Created the Kirkpatrick Model?

Donald Kirkpatrick developed the model in his 1954 PhD dissertation at the University of Wisconsin and published it as a series of journal articles in 1959. The framework has been refined by his family and successors ever since.

Year	Milestone
1954	Donald Kirkpatrick introduces the four levels in his PhD dissertation at the University of Wisconsin
1959	The framework is published as a series of articles in the journal Training and Development
1994	Kirkpatrick's book Evaluating Training Programs: The Four Levels makes the model the de facto industry standard
2010s	Jim Kirkpatrick and Wendy Kayser Kirkpatrick launch the New World Kirkpatrick Model, emphasizing Level 3, business partnership, and planning from Level 4 backward
2016	Jim and Wendy publish Kirkpatrick's Four Levels of Training Evaluation with ATD Press, the definitive New World reference
2026	Vanessa Milara Alzate expands the model to include the performance environment and extends it beyond L&D toward enterprise performance

The sections below walk through each level in order from 1 to 4, which makes them easier to understand. When you actually plan an evaluation, you'll work in the opposite direction, and we'll cover that workflow after the levels.

What Is Level 1: Reaction?

Level 1 (Reaction) measures how participants respond to the training: how engaging, relevant, and useful they found it. This is the most commonly collected evaluation data, usually gathered through a short post-training survey (sometimes called a "smile sheet").

One important update from the New World model: relevance predicts behavior change far better than satisfaction or engagement. A learner can enjoy a workshop and still not apply any of it. But if they tell you the content was directly relevant to their job, they're far more likely to actually use it. Kirkpatrick Partners' current guidance is explicit on this point: relevancy is the Level 1 measure that best indicates whether application and transfer will occur. Build your Level 1 surveys around relevance, not just enjoyment.

Level 1 evaluation also shouldn't wait until the end of a program. Formative pulse checks during the training, like quick polls, check-in questions, or facilitator observations, let you adjust in the moment instead of finding out a week later that learners were lost in module two.

How to Measure Level 1

Post-training surveys (delivered via email, LMS, or in-session form)
In-the-moment polls and pulse checks
Short interviews or focus groups with a sample of participants
Facilitator observations during live sessions

Sample Level 1 Questions

Use these as a starting point for your own questionnaire. Most are rated on a 1 to 5 Likert scale; the last two are open-ended.

The training was directly relevant to my job.
I will be able to apply what I learned in my work.
The content was clear and well-organized.
The instructor (or eLearning experience) kept me engaged.
The pace of the training was appropriate.
The practice activities helped me build confidence.
I would recommend this training to a colleague.
I feel confident I can use what I learned after this session.
What part of the training was most useful to you?
What would you change to make it more useful?

A technical support call center rolls out new screen-sharing software and runs a one-hour webinar teaching agents when to use it, how to initiate a session, and how to handle legal disclaimers. At the end, agents complete a short online survey rating relevance, clarity, and confidence, plus two open-ended questions on what worked and what didn't. The training team uses the results to flag the disclaimer section as confusing and revise it before the next cohort.

What Is Level 2: Learning?

Level 2 (Learning) measures whether learners actually acquired the knowledge, skills, and attitudes the training was designed to teach. This is the cornerstone of most instructional design work. It's where quizzes, demonstrations, and final assessments live.

The New World Kirkpatrick Model expands Level 2 beyond the original "KSA" (knowledge, skills, attitudes) trio. It adds two new indicators: confidence (do learners believe they can do this back on the job?) and commitment (do they intend to actually do it?). Both are strong predictors of whether Level 3 behavior change will happen, and both can be measured with a few well-placed survey questions.

Pre-tests are also worth the effort when feasible. Measuring knowledge before and after training is the cleanest way to attribute gains to the training itself rather than to existing experience.

How to Measure Level 2

Multiple-choice quizzes and written tests (best for knowledge and cognitive skills)
Skill demonstrations, role plays, and simulations (best for procedural and interpersonal skills)
Pre- and post-assessments to measure gain
Confidence and commitment surveys (e.g., "How confident are you that you can do X on Monday?")
Case-study analyses and scenario-based questions

Role plays and demonstrations are the gold standard for measuring skills, but they historically didn't scale: someone had to sit with each learner, watch the performance, and score it by hand. AI conversation simulations have changed this. Learners practice a voice or text conversation with an AI character, and every attempt produces a full transcript plus a score against the rubric you defined. Full disclosure: this is exactly what my own tool, devlin.ai, is built for. You describe the scenario in plain English, get a working simulation, embed it in Storyline, Rise, or your LMS, and the transcripts and scores report back through SCORM or xAPI.

A question I hear from teams running these simulations: do the scores count as Level 2 or Level 3 evidence? Level 2. The practice is still happening inside a learning experience. Level 3 is when you look at and evaluate transcripts of real conversations, once people are doing the work with actual clients and customers. And remember that a simulation on its own, where the learner practices a tough conversation and gets some brief feedback, is valuable but only half the loop. The other half is what happens back on the job, which is exactly where Level 3 picks up.

Sample Level 2 Questions & Assessment Formats

Knowledge check (multiple choice):

Which of the following is the correct first step in initiating a screen-sharing session? (A/B/C/D)
Before sharing your screen with a customer, you must: (select all that apply)

Skill demonstration prompts:

"Walk me through initiating a screen-sharing session with this practice customer."
"Show me how you would handle a customer who declines the screen-sharing request."

Confidence and commitment (1 to 5 scale):

I feel confident I can initiate a screen-sharing session correctly on my next live call.
I plan to use screen sharing on appropriate calls starting this week.

After the call center webinar, agents complete a 10-question multiple-choice quiz on the screen-sharing process and legal disclaimers. They must score 80% or higher to receive certification. They also complete a live role play with their supervisor, initiating a session, walking through the disclaimer, and screen sharing successfully, before they're authorized to use the tool on real customer calls.

A quick contrast: for a coffee-roastery cleaning workshop, written tests don't cut it. Physical procedural skills are best measured by direct observation, like watching each operator clean the machine end to end.

What Is Level 3: Behavior?

Level 3 (Behavior) measures whether people are applying what they learned on the job. Learning something in a classroom and applying it back at your desk are two different things, and Level 3 is where the model starts producing data you can actually act on.

Two practical notes. First, measurement should begin within a few weeks of training, not after 90 days. You'll still find guides that tell you to wait 3 to 6 months before measuring Level 3. This advice is outdated: Kirkpatrick Partners' current guidance is that Level 3 measurement cannot begin at 90 days, because with too much time between learning and measurement "you risk losing the opportunity to adjust" the program. A common cadence is to start observing 2 to 4 weeks post-training and continue through the following months.

Second, behavior change doesn't happen because of training alone. Don Kirkpatrick identified four conditions necessary for behavior change: desire to change, knowledge of what to do, the right climate (a supportive manager, removed obstacles), and rewards for doing it. This framing has held up well in the academic literature; Rouse's 2011 review in Perspectives in Health Information Management walks through these conditions and the transfer barriers that block them. The New World model formalizes the idea as required drivers: the reinforcement, accountability, and support systems that turn learning into behavior, including performance support tools like job aids that carry the learning into the workflow.

How to Measure Level 3

Direct on-the-job observation by supervisors or peers
Performance metrics already tracked by business systems (call data, sales activity, ticket resolution times)
360-degree feedback from managers, peers, and direct reports
Self-reported behavior surveys at intervals (30, 60, 90 days)
Action plans and follow-up coaching conversations
xAPI (Experience API / Tin Can) data for tracking informal and on-the-job activity

Sample Level 3 Questions & Observation Methods

Self-report (sent 30 and 60 days post-training):

How often have you used screen sharing on customer calls in the past 30 days?
What's been the biggest obstacle to applying what you learned?
What support from your manager or team would help you apply this more consistently?

Supervisor observation checklist (sample items):

Agent identified an appropriate moment to offer screen sharing
Agent read the disclaimer accurately and in full
Agent troubleshot connection issues without disengaging the customer

The screen-sharing software is integrated with the call center's performance management platform, so every screen-share session is logged automatically. Three weeks after training, the team pulls a report: what percentage of eligible calls included a screen share? Agents below a threshold get a coaching conversation, not a reprimand. The goal is to find out what's blocking transfer (forgot the steps? worried about customer reaction? no manager reinforcement?) and fix it.

What Is Level 4: Results?

Level 4 (Results) measures whether the training is moving the needle on the outcomes the business cares about: sales, customer satisfaction, retention, safety incidents, output, error rates. This is where training proves its worth, and it's the level most organizations skip. It's also worth asking this question before you build the program at all, which is the territory covered in whether training is the right solution in the first place.

One concept that makes Level 4 more practical is the distinction between leading and lagging indicators. Lagging indicators (quarterly revenue, annual turnover) tell you what already happened. Leading indicators (number of qualified demos booked, first-call resolution rate) predict where the lagging numbers are headed and let you course-correct earlier.

The other big shift is away from chasing a precise ROI number. Calculating the dollar return on a training program with real certainty is extremely hard; there are too many confounding variables. The current Kirkpatrick Model instead presents three complementary measures that together form what Kirkpatrick Partners calls an integrated return:

Return on Expectations (ROE): Did the training deliver what stakeholders said success would look like? Success criteria are defined collaboratively with stakeholders at the start of the project, not imposed afterward.
Return on Performance (ROP): The evidence layer that connects expectations to reality. ROP examines whether people actually performed the critical behaviors on the job (Level 3) and whether those behaviors contributed to the targeted outcomes (Level 4), which supports continuous improvement along the way.
Contributive ROI (cROI): When stakeholders need a financial number, cROI acknowledges that training contributes to business results alongside leadership, systems, culture, and timing, rather than claiming sole credit for them.

Phillips' ROI Methodology (covered in the alternatives section below) takes a different approach and tries to isolate training's financial impact directly. Both views are valid; choose based on what your stakeholders find credible.

How to Measure Level 4

Tracking organization-level KPIs against pre-training baselines
Control-group comparisons where feasible
ROE check-ins with sponsors against pre-defined success criteria
Customer feedback, NPS, and CSAT data
Operational metrics (output, error rates, safety incidents, turnover)

Sample Level 4 Metrics & Questions

For the screen-sharing initiative:

CSAT (customer satisfaction) on calls that included screen sharing vs. calls that did not
First-call resolution rate
Average handle time on issues where screen sharing was used

For stakeholder ROE conversations:

Did we hit the customer-satisfaction targets you defined at the start of the project?
What evidence would convince you this program is worth continuing?
What unintended outcomes, positive or negative, have you observed?

Level 4 Example: Customer Satisfaction Impact

Six months after the screen-sharing rollout, the team compares CSAT scores on calls that used screen sharing against a matched sample that didn't. Calls with screen sharing show a measurable lift in CSAT and a reduction in average handle time. Combined with the Level 3 data showing strong adoption, the training team makes a credible Contributive ROI argument: the program is one of several factors driving the CSAT improvement, and the evidence supports continued investment.

What Is the New World Kirkpatrick Model?

The New World Kirkpatrick Model, developed by Jim Kirkpatrick and Wendy Kayser Kirkpatrick, is the current version of the framework. It keeps the original four levels but fixes the biggest weakness of the 1950s version: the assumption that good training automatically leads to behavior change and business results.

The key additions:

Plan from Level 4 backward. Every project starts by defining business results, then identifies the critical behaviors that produce those results, then the learning required for those behaviors, then the experience that delivers the learning.
Required drivers at Level 3. Reinforcement, encouragement, rewards, and accountability systems that have to exist in the work environment for behavior change to happen. Without them, even excellent training fails to transfer.
Confidence and commitment at Level 2. Two learning indicators that predict transfer better than knowledge alone.
ROE, ROP, and Contributive ROI at Level 4. Stakeholder-defined and contribution-based measures of success that replace the often-unreliable hunt for a precise ROI percentage.

The model took another significant step in 2026. Vanessa Milara Alzate expanded it to include the performance environment as an explicit component: the systems, culture, and external factors that determine whether learning transfers to behavior and then to results. On the current Kirkpatrick Partners model page, Level 3 now includes critical behaviors, performance support, and the performance environment as named components. Alzate also extended the framework's usefulness beyond L&D, applying it across the entire enterprise; Kirkpatrick Partners describes this direction in The New Kirkpatrick Model: From Training Evaluation to Enterprise Performance. The practical takeaway for instructional designers: the model's center of gravity keeps moving away from "did people like the course?" and toward "does the whole system support performance?"

If you're learning the model today, learn the New World version. The original 1959 framework is foundational, but the updated model reflects what practitioners have figured out over six decades of trying to make it work in real organizations.

How Do You Use the Kirkpatrick Model?

You use the Kirkpatrick Model by planning backward: start at Level 4 with the business result, work down to the behaviors and learning that produce it, and build your evaluation instruments before the training launches. Here's the workflow most experienced evaluators follow.

Define Level 4 outcomes with stakeholders
What measurable business result is this training supposed to support? Get a specific target where possible ("sell 800,000 units in year one," "reduce safety incidents by 20%," "improve CSAT by 5 points") and document it as your Return on Expectations. Demand hard numbers, not anecdotes: people having a harder time on calls isn't the same as conversion data showing new hires underperforming against a benchmark. And stay open to the possibility that the best solution isn't training at all; sometimes it's a tool that removes the complexity so people can focus on the work that matters. Run a proper needs assessment before you commit to training.
Identify Level 3 critical behaviors
Working with subject matter experts and frontline managers, list the specific on-the-job behaviors that drive the Level 4 result. Be ruthless and focus on the few behaviors that matter most. Action mapping is built for exactly this critical-behaviors step, and Cathy Moore's Map It approach is the definitive treatment of it. Pair this with a training needs assessment that ties the request to a business metric.
Plan the required drivers
For each critical behavior, identify the reinforcement, accountability, and support that needs to exist post-training. Who reinforces it? What's measured? What's rewarded? Without this step, training won't transfer.
Define Level 2 learning objectives
What knowledge, skills, attitudes, confidence, and commitment do learners need to perform the critical behaviors? Write objectives that point directly at Level 3 behaviors. This is where the analysis work that produces solid learning objectives pays off.
Design the Level 1 experience
Build the training to be relevant first, engaging second. Make sure learners can see the link between what they're doing in the session and what they'll do back on the job.
Build evaluation instruments before launch
Write your surveys, quizzes, observation checklists, and metric dashboards before the training rolls out. If you wait until afterward, you'll miss baseline data and lose credibility with stakeholders.
Implement with formative pulse checks
Don't wait for end-of-training feedback. Check in during the program, after each module, and at intervals post-training (30, 60, 90 days).
Measure, report, and iterate
Share results with sponsors against the ROE you defined in step 1. Where the program is working, scale it. Where it's not, find out whether the issue is design, delivery, or missing required drivers, and then fix the root cause.

Step 1 is where most evaluation efforts live or die, and how far you get depends on who's in the room. In my early client projects, I would try to have this conversation, and the clients would keep bringing the metrics back to the number of courses they needed. Now that I work on more strategic projects and interface directly with Directors and VPs, we can have the Level 4 conversations about which metrics we need to impact at an organizational level. If you're stuck talking course counts, the fix usually isn't a better argument. It's getting closer to the people who own the business metrics.

This workflow pairs well with broader ID frameworks like ADDIE and other instructional design models. Kirkpatrick handles the evaluation strategy; ADDIE handles the design and development workflow.

Ready to put the workflow to work? The builder below walks you through all four levels for your specific program and hands you an evaluation plan you can copy straight into your project doc.

Interactive tool

Build your evaluation plan

Pick your program type, tell me what the business needs, and choose the instruments that fit. You'll leave with a four-level evaluation plan you can copy into your project doc.

Fair warning before you start: most teams only ever measure Levels 1 and 2, because Levels 3 and 4 are harder. But the data that matters most lives at Levels 3 and 4, and behavior change depends more on manager reinforcement than on training quality. If you can't commit to the follow-through, trim the plan rather than pretend.

What kind of training is it?

Who is the audience? (optional)What business outcome does your stakeholder care about? (optional, but do it)

Prefer to read? The full measurement menu behind this tool

Program type	Level	What to measure	Instrument	Timing	Effort
Compliance training	1. Reaction	Perceived relevance to the learner's actual job	Post-training survey built around relevance, not enjoyment	Immediately after training	low
Compliance training	1. Reaction	Confusion points in the content	In-session pulse checks after each policy section	During training	low
Compliance training	1. Reaction	What learners would change	Two open-ended questions at the end of the survey	Immediately after training	low
Compliance training	2. Learning	Knowledge of the policy and required steps	Scored quiz with a pass threshold, scenario questions over recall questions	End of training	medium
Compliance training	2. Learning	Ability to apply the policy in context	Short case analyses graded against a rubric	End of training	medium
Compliance training	2. Learning	Confidence and commitment to comply	Two survey items: can you do this Monday, and do you intend to	End of training	low
Compliance training	3. Behavior	Compliant behavior in real work	Supervisor spot-check observations with a short checklist	Starting 2 to 4 weeks post-training, ongoing	high
Compliance training	3. Behavior	Process adherence captured by systems	Existing system logs and exception reports	Monthly pulls from week 4 onward	medium
Compliance training	3. Behavior	Barriers to compliance	30 and 60 day self-report survey on obstacles	30 and 60 days post-training	low
Compliance training	3. Behavior	Manager reinforcement of the policy	Required-drivers audit: is anyone reinforcing, measuring, or rewarding this	4 weeks post-training	medium
Compliance training	4. Results	Violation and incident rate	Incident reports and audit findings vs pre-training baseline	Quarterly, 3 to 12 months post-training	medium
Compliance training	4. Results	Stakeholder-defined success	Return on Expectations check-in with legal or compliance sponsor	Defined at kickoff, reviewed quarterly	low
Compliance training	4. Results	Audit readiness	Internal audit scores against pre-training baseline	Next audit cycle	medium
New-hire onboarding	1. Reaction	Relevance of week-one content to the actual role	End-of-week-one survey focused on relevance and clarity	End of first week	low
New-hire onboarding	1. Reaction	Pacing and overload	Quick pulse check after each onboarding module	During onboarding	low
New-hire onboarding	1. Reaction	What's missing from the program	Short interviews with a sample of recent hires	End of onboarding	medium
New-hire onboarding	2. Learning	Role knowledge and core procedures	Knowledge check aligned to the role's first real tasks	End of each module	medium
New-hire onboarding	2. Learning	Ability to perform core tasks	Supervised task walkthrough or role play before going live	End of onboarding	high
New-hire onboarding	2. Learning	Confidence and commitment for the first solo week	Confidence and commitment survey items	End of onboarding	low
New-hire onboarding	3. Behavior	Independent performance of core tasks	Manager observation checklist on the job	2 to 4 weeks after onboarding ends	high
New-hire onboarding	3. Behavior	Ramp progress on real work	Existing performance systems: tickets closed, calls handled, tasks shipped	Weekly from week 2 through month 3	medium
New-hire onboarding	3. Behavior	Obstacles new hires hit applying the training	30/60/90 day self-report survey	30, 60, and 90 days	low
New-hire onboarding	3. Behavior	Manager and buddy reinforcement	Required-drivers check: is the manager coaching to the same standards onboarding taught	30 days post-onboarding	medium
New-hire onboarding	4. Results	Time to full productivity	Ramp metric vs pre-program baseline	Per cohort, 3 to 6 months	medium
New-hire onboarding	4. Results	Early retention	90-day and 1-year retention vs baseline cohorts	Quarterly	low
New-hire onboarding	4. Results	Early quality and error rates	QA scores or error rates for new hires' first 90 days	Monthly, first two quarters	medium
Sales enablement	1. Reaction	Relevance to live deals	Post-session survey asking reps if this maps to their current pipeline	Immediately after training	low
Sales enablement	1. Reaction	Engagement during rollout	In-session polls and facilitator observation	During training	low
Sales enablement	1. Reaction	Field credibility of the content	Focus group with a few senior reps	Within a week of training	medium
Sales enablement	2. Learning	Pitch and objection-handling skill	Recorded role play scored against a rubric	End of training	high
Sales enablement	2. Learning	Product and process knowledge	Short scenario-based quiz	End of training	low
Sales enablement	2. Learning	Commitment to use the new approach	Commitment survey item plus a named next deal to try it on	End of training	low
Sales enablement	3. Behavior	Use of the new approach in real deals	CRM activity data: talk tracks logged, plays run, demos booked	Weekly from week 2 onward	medium
Sales enablement	3. Behavior	Quality of application on live calls	Call recording reviews or manager ride-alongs with a checklist	Starting 2 to 4 weeks post-training	high
Sales enablement	3. Behavior	Manager reinforcement in pipeline reviews	Required-drivers check: are managers coaching to the new approach	4 weeks post-training	medium
Sales enablement	4. Results	Leading indicators of revenue	Qualified demos booked, proposals out, stage-conversion rates	Monthly from month 1	medium
Sales enablement	4. Results	Win rate and quota attainment	Lagging sales KPIs vs pre-training baseline	Quarterly, 3 to 12 months	low
Sales enablement	4. Results	Stakeholder-defined success	Return on Expectations check-in with the sales VP	Defined at kickoff, reviewed quarterly	low
Technical skills	1. Reaction	Relevance and pacing for mixed skill levels	Post-training survey with relevance and pace items	Immediately after training	low
Technical skills	1. Reaction	Confusion points in hands-on sections	Pulse checks after each lab or practice block	During training	low
Technical skills	1. Reaction	Perceived usefulness of the practice environment	Two open-ended survey questions on the labs	Immediately after training	low
Technical skills	2. Learning	Hands-on skill with the tool or procedure	Skill demonstration: perform the task end to end, observed	End of training	high
Technical skills	2. Learning	Knowledge gain over baseline	Pre- and post-assessment with matched items	Before training and at the end	medium
Technical skills	2. Learning	Confidence to use the skill unsupervised	Confidence survey items after the demonstration	End of training	low
Technical skills	3. Behavior	Actual tool usage on the job	System logs and usage analytics, xAPI if you have it	Weekly from week 2 onward	medium
Technical skills	3. Behavior	Quality of applied work	QA reviews or peer observation with a checklist	Starting 2 to 4 weeks post-training	high
Technical skills	3. Behavior	Obstacles to applying the skill	30 and 60 day self-report on blockers	30 and 60 days post-training	low
Technical skills	4. Results	Operational metrics the skill should move	Resolution time, error rate, or output vs baseline	Monthly, 3 to 12 months	medium
Technical skills	4. Results	Downstream customer impact	CSAT or NPS on work that used the new skill vs work that didn't	3 to 6 months post-training	high
Technical skills	4. Results	Stakeholder-defined success	Return on Expectations check-in against kickoff criteria	Defined at kickoff, reviewed at 6 months	low
Leadership development	1. Reaction	Relevance to each leader's real situation	Post-session survey centered on relevance items	After each session	low
Leadership development	1. Reaction	Session-by-session engagement	Cohort check-ins and facilitator observation	During the program	low
Leadership development	1. Reaction	Perceived applicability of each model taught	One open-ended question per session: what will you use this week	After each session	low
Leadership development	2. Learning	Judgment in realistic scenarios	Case-study analysis scored against a rubric	End of program	medium
Leadership development	2. Learning	Skill in coaching conversations	Observed role play of a real conversation they need to have	During and end of program	high
Leadership development	2. Learning	Confidence and commitment to lead differently	Confidence and commitment items plus a written action plan	End of program	low
Leadership development	3. Behavior	Changed behavior as their team experiences it	360-degree feedback vs pre-program baseline	Baseline before, repeat at 3 to 6 months	high
Leadership development	3. Behavior	Follow-through on action plans	Structured coaching conversations reviewing the plan	30, 60, and 90 days	medium
Leadership development	3. Behavior	Direct-report experience over time	Short pulse survey to each leader's team	60 and 120 days post-program	medium
Leadership development	4. Results	Team retention and engagement	Retention and engagement scores on participants' teams vs comparable teams	6 to 12 months	medium
Leadership development	4. Results	Internal mobility of participants	Promotion and readiness data vs baseline	12 months	low
Leadership development	4. Results	Stakeholder-defined success	Return on Expectations review with the executive sponsor	Defined at kickoff, reviewed at 6 and 12 months	low
Something else	1. Reaction	Perceived relevance to the job	Post-training survey built around relevance, not satisfaction	Immediately after training	low
Something else	1. Reaction	Confusion and drop-off points	In-the-moment pulse checks during the program	During training	low
Something else	1. Reaction	What worked and what to change	Short interviews or focus group with a sample of learners	Within a week of training	medium
Something else	2. Learning	Knowledge and skill gain over baseline	Pre- and post-assessment with matched items	Before training and at the end	medium
Something else	2. Learning	Ability to perform the target skill	Skill demonstration or scenario assessment, observed	End of training	high
Something else	2. Learning	Confidence and commitment to apply it	Two survey items: can you do this on Monday, and will you	End of training	low
Something else	3. Behavior	On-the-job application of the training	Direct observation by supervisors or peers with a short checklist	Starting 2 to 4 weeks post-training	high
Something else	3. Behavior	Behavior captured by existing systems	Performance data the business already tracks	Monthly from week 4	medium
Something else	3. Behavior	Obstacles to applying it	30/60/90 day self-report survey	30, 60, and 90 days	low
Something else	3. Behavior	Reinforcement in the work environment	Required-drivers audit: who reinforces, what's measured, what's rewarded	4 weeks post-training	medium
Something else	4. Results	The business KPI this training supports	KPI tracking vs pre-training baseline	3 to 12 months, ongoing	medium
Something else	4. Results	Early signals the KPI will move	Leading indicators reviewed monthly	Monthly from month 1	medium
Something else	4. Results	Stakeholder-defined success	Return on Expectations check-in against criteria set at kickoff	Defined at kickoff, reviewed quarterly	low
Something else	4. Results	Effect vs a comparison group	Control-group comparison where feasible	3 to 6 months post-training	high

Kirkpatrick Model Template and Sample Questionnaire

Use this template to plan your evaluation with the Kirkpatrick Model: for each level, define what success looks like, how you'll measure it, and when. It's the same evaluation-planning structure we teach at Peck Academy, my licensed career school for people transitioning into instructional design.

Level 1: Reaction

Success Criteria	Planned Method(s)	Timing
[Insert success criteria here]	[Insert proposed method(s) here]	[Insert proposed timing here]

Level 2: Learning

Success Criteria	Planned Method(s)	Timing
[Insert success criteria here]	[Insert proposed method(s) here]	[Insert proposed timing here]

Level 3: Behavior

Success Criteria	Planned Method(s)	Timing
[Insert success criteria here]	[Insert proposed method(s) here]	[Insert proposed timing here]

Level 4: Results

Success Criteria	Planned Method(s)	Timing
[Insert success criteria here]	[Insert proposed method(s) here]	[Insert proposed timing here]

Sample Questionnaire by Level

The sample questions earlier in this guide can be assembled into a complete questionnaire. A practical approach:

End-of-training survey (Level 1): 8 to 10 items covering relevance, clarity, engagement, confidence, and two open-ended questions.
Knowledge assessment (Level 2): 10 to 15 multiple-choice items aligned to learning objectives, plus 2 or 3 confidence and commitment items on a 5-point scale.
30/60/90-day follow-up (Level 3): 5 to 8 self-report items on frequency of application, perceived obstacles, and support received.
Stakeholder ROE check-in (Level 4): 3 to 5 questions asking sponsors to evaluate progress against the success criteria defined at kickoff.

What Are the Criticisms of the Kirkpatrick Model?

The main criticisms of the Kirkpatrick Model are that the causal links between levels are weaker than the model implies, that organizations rarely measure Levels 3 and 4, that the model doesn't isolate training from environmental factors, and that Level 4 attribution is difficult to defend. Knowing these limitations makes you better at defending your evaluation choices to a skeptical stakeholder.

The causal-link assumption is weak. The model implies that positive Level 1 reactions lead to Level 2 learning, which leads to Level 3 behavior, which leads to Level 4 results. Decades of research show these relationships are far weaker than the model suggests. Will Thalheimer's critique, which led him to build the LTEM alternative covered below, is the sharpest version of this argument: strong Level 1 scores tell you almost nothing about whether Level 3 transfer will happen.
Most organizations stop at Levels 1 and 2. Because Levels 3 and 4 are harder to measure, they get skipped. According to ATD's 2016 report Evaluating Learning: Getting to Measurements That Matter, only 35% of the 199 organizations surveyed evaluated the business results of their learning programs. The data that matters most is the data least often collected.
The model doesn't isolate training from the environment. If Level 3 behavior doesn't change, was the training bad or did the work environment fail to support it? The original model doesn't answer this clearly. The New World model's required drivers and, as of 2026, the performance environment address this directly, and it's also why the problem often isn't a lack of training at all.
Attribution at Level 4 is hard. Business results are influenced by economic conditions, product changes, leadership, market timing, and dozens of other variables. Claiming training caused a specific outcome is rarely defensible. ROE, ROP, and Contributive ROI are more credible framings.
The model is descriptive, not prescriptive. It tells you what to measure but not how to design effective training. It needs to be paired with instructional design models and learning science.

None of this invalidates the framework. It's still the most useful starting point we have. But pretending the model is airtight does the profession no favors.

What Are the Alternatives to the Kirkpatrick Model?

The main alternatives are the Phillips ROI Methodology, LTEM, the CIRO Model, Anderson's Value of Learning Model, and Brinkerhoff's Success Case Method. Each was developed to address a Kirkpatrick limitation or take a different angle on the same problem.

Model	What It Adds	Best For
Phillips ROI Methodology	Adds a fifth level (financial ROI) and provides a specific methodology for isolating training's impact and converting results to monetary value	Organizations that need a defensible dollar-figure ROI
LTEM (Thalheimer)	An eight-tier Learning-Transfer Evaluation Model built specifically to fix Kirkpatrick's weak causal links, with tiers that distinguish attendance and activity from decision-making competence and transfer	Teams that want evaluation grounded in learning research rather than reaction data
CIRO Model (Warr, Bird, Rackham)	Four stages: Context, Input, Reaction, Output. Adds front-end analysis (context and input) that Kirkpatrick assumes you've already done	Programs that need formal needs analysis built into the evaluation framework
Anderson's Value of Learning Model	Three-stage model emphasizing strategic alignment of learning with business priorities before measurement	Senior L&D leaders aligning a training portfolio with strategy
Brinkerhoff's Success Case Method	Identifies the most and least successful cases and studies them in depth, rather than averaging across all participants	Quickly understanding what makes training transfer (or fail to transfer) in real conditions

In practice, most organizations end up using Kirkpatrick as the spine and borrowing from Phillips (for ROI), LTEM (for rigor about what learning data actually means), CIRO (for front-end context), or Brinkerhoff (for case-based insight) when the situation calls for it.

Frequently asked questions

What are the four levels of the Kirkpatrick Model?

Level 1 (Reaction) measures how learners respond to the training. Level 2 (Learning) measures what they know and can do. Level 3 (Behavior) measures whether they apply it on the job. Level 4 (Results) measures the impact on business outcomes.

What is the Kirkpatrick Model used for?

The Kirkpatrick Model is used to evaluate whether training programs work: whether learners found them relevant, learned from them, changed their on-the-job behavior, and produced business results. Practitioners also use it in reverse as a planning tool, defining the Level 4 business outcome first and designing the training backward from it.

Who developed the Kirkpatrick Model and when?

Donald Kirkpatrick developed the model as part of his 1954 PhD dissertation at the University of Wisconsin and published it through a series of articles in 1959. His son Jim Kirkpatrick and daughter-in-law Wendy Kayser Kirkpatrick updated it into the New World Kirkpatrick Model in the 2010s, and Vanessa Milara Alzate expanded it in 2026 to include the performance environment.

When should each level be measured?

Level 1: during and immediately after training. Level 2: at the end of training (and pre-training as a baseline where possible). Level 3: starting within 2 to 4 weeks of training, not after 90 days, and continuing over the following months. Level 4: 3 to 12 months post-training, with ongoing tracking.

What's the difference between the original and New World Kirkpatrick Model?

The New World model adds required drivers (the post-training reinforcement that enables behavior change), confidence and commitment as Level 2 indicators, an emphasis on planning from Level 4 backward, and ROE, ROP, and Contributive ROI as more practical alternatives to financial ROI. The 2026 expansion adds the performance environment as an explicit Level 3 component and extends the model beyond L&D toward enterprise performance.

What is the difference between ADDIE and the Kirkpatrick Model?

ADDIE is an instructional design process (Analysis, Design, Development, Implementation, Evaluation) that guides how you build training. The Kirkpatrick Model is an evaluation framework that measures whether the training worked. They pair naturally: Kirkpatrick supplies the evaluation strategy inside ADDIE's evaluation phase. See instructional design models and theories for how the frameworks fit together.

Is the Kirkpatrick Model still relevant?

Yes. ATD's 2025 research found it is the most common evaluation framework, used by 71% of organizations surveyed. The New World version addresses most of the modern critiques of the original, and the four-level vocabulary is universal among practitioners.

What are the main criticisms of the Kirkpatrick Model?

The main criticisms are that the causal links between levels are weaker than the model implies, that organizations rarely measure Levels 3 and 4 in practice (ATD's 2016 research found only 35% evaluate business results), that the original model doesn't account for environmental and transfer factors, and that Level 4 attribution is difficult to defend.

How does Kirkpatrick compare to the Phillips ROI Model?

Phillips adds a fifth level (financial ROI) and a specific methodology for isolating training's monetary impact. Kirkpatrick (especially the New World version) prefers Return on Expectations, Return on Performance, and Contributive ROI, which acknowledge that training contributes to outcomes alongside other factors rather than claiming sole financial credit.

Putting the Kirkpatrick Model into Practice

The Kirkpatrick Model isn't a checklist. It's a discipline. Used well, it forces you to answer two hard questions before you build a single slide: what does the business actually need? And what will people need to do differently for that to happen? Everything else follows from there.

If you only do one thing with this framework, plan from Level 4 backward. This single habit separates evaluation that informs decisions from evaluation that fills out a checkbox. And it points at where I think the whole field is headed:

We're moving from content delivery to practice delivery. And practice is where performance actually changes.Devlin Peck

Practice is also where Levels 2 and 3 finally become measurable instead of aspirational. If this way of working appeals to you, where the business result comes first and the course is just one possible means to it, you're already thinking like a performance consultant. Here's what operating as a performance consultant actually looks like.