Abstract Education and training interventions can be evaluated through the success of learning outcomes. Kirkpatrick's four-level model is a widely accepted and highly popular evaluation tool. However, some criticise the model's shortcomings. This article will examine the extent to which the four-level model can evaluate design and technology students' learning about aesthetics after an intervention by reporting our use of an augmented version of the four-level model. We examine the results in terms of students' reaction to the intervention, their long-term learning and their behaviour changes by studying their visual analyses and drawings through segment codes. We found that, in order to uncover the obscurities imbedded in aesthetics and to explicate the complexities, we could not use the four-level-model on its own, but had to revert to a more augmented version.