This research investigates the linkage between training evaluation, learning design and training transfer. A new training evaluation model, (i.e., learning-transfer evaluation model [LTEM]), was used to examine its ability to provide evaluative evidence through robust assessments in pre-, post- and delayed assessments. The model was used to improve the training design of a nursing training programme and to nudge the stakeholders to change their training practices. A mixed-methods research approach, called convergent parallel, was employed to study the influence of the LTEM model in training design improvement by assessing the perceptions of nurse trainees, trainers and a nurse educator towards the instructional design of the training programme. A quantitative approach was used to examine the extent to which the improved programme using an iterative design cycle would result in improved performance among four groups (i.e., one baseline and three treatment groups). The integrated data showed that the LTEM model influenced the stakeholders to focus more on skills practice. With this shift of mindset, the training design was enhanced based on the principles of learning design, cognitive science and multimedia principles. Results of the pre-, post- and delayed posttests showed higher mean scores in the three treatment groups compared to the baseline group. For the measure of transfer, while the treatment groups scored higher, there was no statistically significant difference among all four groups. This could be attributed to nurses' varied levels of experience, negative transfer and the use of different evaluation matrices in the study and by the hospital.