Kirkpatrick Four-Level Training Evaluation

Description

Evaluate a training or learning intervention at four levels: trainee reactions, learning acquired, behaviour change in real use, and organisational results.

The Kirkpatrick model is the most widely used frame for evaluating training. Level 1, reactions, asks whether trainees liked the training and found it useful. Level 2, learning, tests whether targeted knowledge and skills were acquired against the objectives. Level 3, behaviour, observes whether performance changed on the job, after time for transfer. Level 4, results, tracks organisational outcomes such as costs, efficiency, quality, turnover and return on investment. Evaluation is planned during needs analysis so each level has criteria, and it should explain why training worked, not just whether. Critics note the levels lack theory and are not necessarily causally linked: high satisfaction does not imply learning, nor learning transfer, so a multilevel view is recommended.

Attributes

Loss type
Siloed
Data
Mixed
Effort
Months

Steps

  1. Before training, define learning objectives from organisational, task and person needs analysis so each level has explicit criteria.
  2. Level 1: collect reactions (liking, perceived usefulness, ratings of materials) at the end of training.
  3. Level 2: test acquisition of the targeted knowledge and skills against the objectives, ideally with a pre-test.
  4. Level 3: observe or measure behaviour in real use after allowing time for transfer, for example 4-12 weeks later.
  5. Level 4: track organisational outcomes such as support costs, efficiency, quality, turnover and ROI.
  6. Compare levels to find where the chain breaks, then redesign the training, job aids or post-training climate.

Sources

  • Handbook of Human Factors and Ergonomics (Salvendy & Karwowski, 2021)

Also known as

  • Kirkpatrick model
  • Reaction-learning-behaviour-results evaluation
  • Four-level training evaluation

Categories

  • Research Operations
  • Research Design & Rigor