Why Learning Evaluation Falls Short — and How LTEM Reframes the Conversation

For decades, Kirkpatrick has shaped how organisations approach training evaluation. It gave the profession a shared language and a useful structure. And to be fair, that structure still matters. But when the real question becomes whether learning actually transfers to the job, many teams hit a wall.

Reaction scores look healthy. Knowledge checks pass comfortably. Dashboards show green. Yet behaviour on the ground tells a more complicated story. Performance shifts in some pockets, stalls in others, and leaders are left wondering what truly changed.

That gap is exactly what the Learning-Transfer Evaluation Model, or LTEM, was built to address. Developed by Will Thalheimer and first published in 2018, LTEM emerged from a simple but uncomfortable observation: not all evidence of learning is equally meaningful. Drawing on research in learning science and cognitive psychology, Thalheimer challenged the field to think more carefully about what our measures actually prove. Some measures reassure us. Others genuinely indicate that people can make sound decisions, perform tasks competently, and deliver results where it counts.

Since its original publication, LTEM has evolved through multiple refinements, now commonly referred to as Version 13. Each iteration sharpened its focus on evidence quality and decision-making. Importantly, LTEM does not dismiss earlier frameworks. It reframes the conversation. It asks a more precise question: what level of evidence do we need to confidently say learning is driving performance?

Within this series, LTEM complements both Kirkpatrick and Phillips. Kirkpatrick structures the thinking. Phillips strengthens the financial lens. LTEM focuses the spotlight on transfer itself, offering a clearer, more practical map of what to measure and when.

LTEM in 60 seconds: An eight-tier framework that helps L&D teams prioritise the kinds of evidence that most credibly demonstrate learning transfer and real-world performance.

At its best, LTEM brings discipline. It does not encourage measuring everything simply because we can. It encourages measuring what matters. As we unpack the eight tiers, you will see how LTEM guides teams away from surface indicators and toward credible signals of competence and impact. That shift, applied thoughtfully, supports better decisions and stronger performance outcomes.

What LTEM is

The Learning-Transfer Evaluation Model is a practical framework for thinking about evidence. It recognises something many of us intuitively know but do not always act on: not all measures of learning carry the same weight.

Some measures tell us that people showed up. Others confirm they understood the material. Far fewer demonstrate that they can make the right decisions under pressure, execute critical tasks to standard, and apply their skills effectively in real work situations.

LTEM organises evaluation evidence into eight tiers. These tiers progress from weaker forms of evidence at the base to stronger, more performance-linked evidence at the top. Attendance and learner perceptions sit near the bottom. Decision-making competence, task execution, and measurable workplace effects sit near the top. The progression is deliberate. As you move upward, the evidence becomes more directly tied to behaviour and organisational results.

The mindset behind LTEM is grounded in learning science. It pushes us to think about how people actually learn and perform, not just how they recall information. That means shifting attention from isolated fact recall to decision quality in context. It means asking whether someone can do the task, not just describe it. In practice, this often reshapes both how programmes are assessed and how they are designed in the first place.

Compared to Kirkpatrick, LTEM adds granularity where it matters most. Kirkpatrick separates learning from behaviour. LTEM zooms in on the space between those two points. It highlights decision-making competence and task competence as the necessary bridges to transfer. That additional clarity makes it particularly valuable for teams serious about evidence-informed evaluation and performance improvement.

The practical payoff is significant. When LTEM is applied with intent, teams avoid spending time and budget on measures that offer limited insight. Instead, they create a focused measurement map that links each learning objective to the level of evidence required to demonstrate real change. A well-executed LTEM evaluation typically results in a clear hierarchy of evidence, aligned instruments, and targeted recommendations that strengthen transfer.

In other words, LTEM helps evaluation earn its place at the performance table.

LTEM’s eight tiers

The power of LTEM lies in its structure. The eight tiers form a ladder of evidence, moving from measures that are easy to collect but weak in meaning to those that are harder to obtain but far more compelling. Walking through them one by one makes the logic clear.

1. Attendance Definition: Who attended or completed the learning. What it tells you: Exposure, not competence. Measurement methods: LMS completion data, attendance registers, sign-in sheets. Example metric: 92 percent of targeted managers completed the programme.

Attendance establishes reach. It confirms participation. It does not tell you whether anything changed.

2. Activity Definition: The extent of learner participation in learning activities. What it tells you: Observable engagement, not capability. Measurement methods: Time on task, clickstream data, participation counts, discussion posts. Example metric: Average time spent on simulation: 18 minutes.

Activity data can signal effort or interest. But effort alone does not equal competence.

3. Learner Perceptions Definition: Learners’ views on relevance, satisfaction, and confidence. What it tells you: Perceived value and willingness to apply. Measurement methods: Post-programme surveys, pulse checks, net promoter-style items. Example instrument: “This programme prepared me to handle real client objections” rated on a five-point scale.

Perceptions matter because motivation affects transfer. Still, positive ratings are not proof of improved performance.

4A. Foundational Knowledge Definition: Knowledge that supports sound decisions. What it tells you: Conceptual understanding that underpins performance. Measurement methods: Constructed response questions, scenario-based items, applied quizzes. Example item: “Given this customer scenario, which regulatory principle applies and why?”

This level tests whether learners understand what sits beneath good performance. It is useful, but it remains indirect.

4B. Trivial Knowledge Definition: Facts that do not meaningfully drive performance decisions. What it tells you: Recall of peripheral information. Measurement methods: Simple fact-recall questions. Example metric: Percentage of learners recalling the year a policy was introduced.

LTEM separates foundational knowledge from trivial recall for a reason. Measuring facts that do not influence decisions can create the illusion of rigour without strengthening performance.

5. Decision-making competence Definition: The ability to choose appropriate actions in realistic scenarios. What it tells you: Whether learners can apply knowledge to make sound judgments. Measurement methods: Branching simulations, case analyses, situational judgement tests. Example item: Present a complex scenario and ask learners to select and justify the best course of action.

Here, we move closer to real work. Decision-making competence begins to approximate the pressures and trade-offs employees face daily.

6. Task competence Definition: The ability to perform key tasks to a defined standard. What it tells you: Observable capability in action. Measurement methods: Skills demonstrations, observed assessments, structured manager checklists, performance rubrics. Example rubric item: “Conducts client needs analysis using approved questioning framework” scored against defined criteria.

This is where talk becomes action. Task competence provides strong evidence that someone can perform as required.

7. Transfer Definition: Successful application of skills in targeted workplace situations. What it tells you: Behaviour change in real context. Measurement methods: Workplace observations, manager corroboration, performance logs, xAPI traces capturing real usage. Example metric: Percentage of trained supervisors consistently applying a new safety protocol during site inspections.

At this tier, learning has left the classroom and entered the workplace.

8. Effects of transfer Definition: Measurable impact resulting from applied skills. What it tells you: Organisational outcomes influenced by transfer. Measurement methods: KPI tracking, productivity metrics, error reduction rates, revenue growth, safety incident data. Example metric: 15 percent reduction in customer complaints following a sales capability intervention.

This is the strongest level of evidence. It is also the most complex, often requiring integrated data and careful interpretation.

Taken together, the tiers create a structured way of thinking about proof. Lower tiers establish reach and engagement. Middle tiers test decision quality and competence. Higher tiers confirm behavioural change and business effects. Combining tiers strengthens credibility. For example, strong Tier 5 scenario results supported by Tier 7 workplace observations form a far more compelling case than either measure alone.

The tiers are not a checklist to complete mechanically. They are a guide. Used thoughtfully, they help teams move beyond surface indicators and build a clearer, more honest picture of learning’s contribution to results.

Steps for implementing LTEM: mapping, measurement, and analysis

Understanding LTEM conceptually is useful. Applying it well is what makes the difference. The goal is not to build a complex evaluation machine. The goal is to generate evidence that helps leaders make better performance decisions.

A practical approach can keep the process focused.

Step 1 — Define evaluation purpose and stakeholders

Start with the decision. What will this evaluation influence? Continue the programme? Refine it? Scale it? Stop it? When the decision is clear, the measurement becomes purposeful.

Bring the right people into the conversation early. This usually includes a business sponsor, the L&D owner, line managers, and a data owner with access to operational metrics. When these voices align, evaluation becomes a shared performance dialogue rather than an isolated L&D exercise.

Step 2 — Map learning outcomes to LTEM tiers

For each critical learning objective, identify the smallest set of tiers needed to demonstrate meaningful change. There is no prize for measuring all eight tiers.

If the goal is to improve frontline sales conversations, knowledge checks alone will not suffice. You may need Tier 5 decision scenarios and Tier 6 observed task competence to demonstrate readiness. Mapping outcomes to tiers creates a visible bridge between learning design and workplace performance.

A simple working template helps: Learning objective → Target LTEM tier(s) → Measurement instrument → Performance indicator.

Step 3 — Select instruments per tier

With tiers defined, choose instruments that match the required evidence strength. Lower tiers may rely on LMS data and short surveys. Middle tiers often require scenario-based assessments. Higher tiers may demand observation tools, performance rubrics, or integrated data streams.

There are always trade-offs. Some measures are quick but shallow. Others are powerful but require systems integration and manager time. The key is proportionality. Select instruments that provide sufficient evidence for the decision at hand.

Step 4 — Sampling and data collection plan

Decide whether to evaluate everyone or a representative sample. Large-scale programmes often benefit from sampling to manage cost and effort. Pilots and high-risk initiatives may justify full-cohort evaluation.

Set realistic timelines. Transfer does not happen overnight. Allow space to observe behaviour in context.

Clarify consent and data use, especially when performance metrics are involved. Manager participation is particularly important for Tier 6 and Tier 7 evidence. Without it, higher-tier claims weaken quickly.

Step 5 — Analysis and triangulation

LTEM encourages looking across tiers rather than at any one in isolation. Lower-tier results may suggest patterns, but higher-tier evidence should validate claims.

Where possible, connect behavioural data to outcomes. A simple transfer rate, such as the proportion of target tasks performed correctly in observed settings, can provide meaningful insight. From there, explore correlations with relevant KPIs.

Attribution requires humility. Performance shifts rarely have a single cause. The aim is to build a credible chain of evidence, not to overstate certainty.

Step 6 — Reporting and action

Evidence only matters if it drives decisions. Reporting should therefore be concise and purposeful. An executive summary can outline key findings per tier. A tiered dashboard can show progression from participation to impact. Narrative examples can add context and clarity.

Most importantly, recommendations should link directly to specific tiers. Weak decision competence calls for better practice scenarios. Low transfer despite strong competence may signal environmental barriers such as workload or insufficient manager reinforcement.

Used this way, LTEM becomes more than a model. It becomes a disciplined way of improving performance outcomes through better evidence and smarter action.

When and why to use LTEM

LTEM is particularly valuable when transfer is the explicit objective. It fits naturally in programmes that develop complex skills, support decision-heavy roles, or operate in high-risk environments such as healthcare, safety, financial services, or frontline sales.

In these contexts, attendance data offers little comfort. Leaders need credible evidence that employees can make sound decisions and execute tasks under real conditions.

LTEM supports this by aligning measurement with how performance actually unfolds. It encourages scenario-based practice, competency assessments, and observation strategies that mirror real work. In doing so, it strengthens both evaluation and design.

LTEM also works well alongside Kirkpatrick and Phillips. Establish a credible learning-to-behaviour link using LTEM. Then, where possible, connect that behavioural evidence to business outcomes through ROI analysis or KPI tracking. The result is a stronger, more defensible performance narrative.

When used intentionally, LTEM helps teams concentrate effort where it will most influence results.

When not to use LTEM — limitations and pitfalls

LTEM is not necessary in every situation. For small compliance modules where completion alone satisfies regulatory requirements, a tiered evaluation may introduce unnecessary complexity. Similarly, if there is no realistic way to observe workplace behaviour, perhaps due to limited manager engagement or lack of data access, progressing beyond lower tiers becomes difficult.

There are also common pitfalls. Measurement overload can creep in when teams attempt to collect data at every tier. Confusing foundational knowledge with trivial recall can distort priorities. Poorly designed scenarios may inflate perceived decision competence. And relying solely on self-report for higher tiers weakens the credibility of claims.

Data integration presents further challenges. Linking learning records to operational KPIs requires coordination, clean systems, and thoughtful interpretation. Multiple variables influence performance outcomes.

The mitigation is simple in principle, though not always easy in practice. Start small. Map tiers pragmatically. Pilot instruments. Secure manager and data owner commitment early. Applied thoughtfully, LTEM remains a powerful guide. Applied mechanically, it risks becoming another reporting routine.

Case examples (preferably African/South African context)

Frameworks become tangible when grounded in real work. The following examples illustrate how LTEM thinking can strengthen transfer and performance outcomes. Where necessary, cases are anonymised.

Example 1: South African retail bank (anonymised) Context: A national retail bank upskilling frontline sales advisors to improve consultative selling and reduce customer churn. Intervention: Blended learning programme incorporating scenario practice and structured coaching. LTEM mapping: Tier 4A scenario-based knowledge checks, Tier 5 decision simulations, Tier 6 observed client conversations scored against a rubric, Tier 7 manager-verified application in branch settings. Key findings: Advisors performed well on knowledge checks but showed variability in live conversations. Decision simulation scores predicted subsequent performance quality. Action taken: Expanded structured coaching, introduced conversation guides as job aids, and aligned manager feedback to the Tier 6 rubric. Outcome: Greater consistency in client needs analysis and measurable uplift in conversion rates within pilot branches.

Example 2: South African education NGO (anonymised) Context: Teacher development initiative focused on improving early grade reading instruction. Intervention: Workshops combined with classroom practice and peer feedback. LTEM mapping: Tier 5 lesson-planning scenarios, Tier 6 observed teaching practice using competency checklists, Tier 7 classroom implementation logs, Tier 8 learner reading improvements. Key findings: Teachers understood the concepts but required structured follow-up to sustain classroom transfer. Action taken: Introduced peer observation cycles and simplified instructional routines. Outcome: Improved fidelity of reading instruction and measurable gains in learner fluency across supported schools.

Example 3: International healthcare setting Context: Hospital system strengthening clinical decision-making under pressure. Intervention: Simulation-based training for nursing staff. LTEM mapping: Tier 5 high-fidelity simulations, Tier 6 assessed task performance in simulated environments, Tier 7 supervisor observation in live settings, Tier 8 reduction in procedural errors. Key findings: Simulation scores correlated with improved protocol adherence in practice. Action taken: Increased simulation frequency and standardised feedback criteria. Outcome: Reduction in avoidable safety incidents over subsequent reporting cycles.

How to anonymise and present LTEM cases to stakeholders Do: Focus on roles, functions, and outcomes rather than naming individuals or specific sites. Do: Pair narrative examples with explicit tier mapping to demonstrate evidence strength. Do: Present both strengths and gaps to maintain credibility. Don’t: Claim direct causality without acknowledging contributing factors. Don’t: Share identifiable performance data without appropriate consent.

Across these examples, the pattern is consistent. When LTEM is applied deliberately, it highlights where competence genuinely develops and where organisational conditions require attention. Evaluation becomes forward-looking rather than retrospective.

Ultimately, LTEM is less about adding another framework and more about sharpening focus. It asks us to seek stronger evidence, design for real decisions, and connect learning directly to meaningful work. When measurement aligns with how performance actually unfolds, learning becomes a deliberate driver of organisational growth.

One clear takeaway: measure for transfer with intention, and performance improvement becomes something you design for, not something you hope for.

Further Reading