Kirkpatrick's Model in Practice: Strengths and Limitations
The standard reference model for training evaluation is genuinely useful — and genuinely limited. Knowing where each applies matters more than the model itself.
The four levels, briefly
Kirkpatrick's model evaluates across four levels: reaction (did people find it valuable), learning (did knowledge or skill increase), behaviour (did on-the-job behaviour change), and results (did the organisational outcome improve). Each level answers a different, progressively harder question.
Where it's genuinely useful
The model's real value is structural — it forces a distinction between reaction and results that's otherwise easy to blur, and stops an organisation mistaking a good reaction score for proof of impact. As a checklist for what to measure and when, it holds up well.
Where it breaks down in practice
The model is far more often applied at level one or two than at level three or four, because reaction and learning are easy to measure immediately, while behaviour and results take longer and require a baseline most organisations never captured. In practice, 'we use Kirkpatrick' frequently means 'we measure reaction' — which defeats the model's actual purpose.
Using it as a starting point, not the whole answer
Kirkpatrick tells you what to measure; it doesn't build the baseline, the timeframe, or the organisational will to measure at levels three and four. Treating it as a complete evaluation methodology, rather than a useful frame that still needs a proper evaluation plan behind it, is where most of the disappointment with the model actually comes from.
Common questions on this topic.
It remains a useful frame, particularly for distinguishing reaction from actual impact — the limitation is in how it's typically applied, not in the model itself.
Not necessarily, but stopping at level one or two and calling it evaluation is the specific mistake to avoid — at minimum, behaviour change should be considered where it's the outcome that actually matters.
Several newer models exist, but most share the same underlying requirement — a baseline and a defined timeframe — so the practical fix is usually building an evaluation plan properly, whichever model frames it.
Want this thinking applied to your organisation?
Insight is useful. Applied insight changes outcomes. Let's talk about yours.