Performance Evaluation: Proven Approaches for Improving Program and Organizational Performance

COMMENTS ON THE MODEL

Holton (1996) criticizes the four-level model as simply a taxonomy with no causal relationships among the various levels. Alliger and Janak (1989) acknowledge the simplicity of Kirkpatrick's four levels but fault his causality attribution, focusing on rejecting three critical assumptions: (1) training directly causes change, (2) each level is causally linked to the next highest level, and (3) correlations between levels are positive. They agree that levels 3 (behavior) and 4 (results) usually suggest a causal relationship, but argue that other variables are just as influential in performance improvement. In particular, they cite the influence of motivation, attitudes, and environmental context (support factors) as alternative ...