Transfer Monitoring from University to Industry

By Clemens Proyer .
Advised by Alexandra Mazak and Christian Huemer

The measurement of the knowledge change of employees as well as the subsequent transfer within their companies is discussed in this thesis. Although these two terms are often used synonymously, there is a clear difference between them. Learning is adapting to a situation whereas transfer is applying this knowledge to similar situations. There are various approaches to measuring learning success or transfer, most of which originate in educational science. In this thesis we consider the special case of innovation courses, where there are various additional requirements that must be met for the measurement. Unfortunately, the existing frameworks are not designed for these requirements and are therefore not sufficient. An innovation course is a long-term course in which employees of various companies are taught and trained in a certain topic. Such an innovation course consists of several modules for which both the measurement of learning success and knowledge transfer for the individual participants must take place in the various modules. To achieve this and to make the measurements repeatable and objective, we have developed a framework for transfer monitoring from university to industry. We use the Design Science Approach to develop the framework. However, the goal is not to create a static artefact that can only be applied to the course of our case study, but to design a framework that is also easily adaptable and applicable in other innovation courses or in a similar environment. To test and improve this framework, we use it in four modules of the DigiTrans 4.0 innovation course. For three of the four modules of our case study, the difference between the knowledge before the module and at the end is statistically significant. We also create linear models to explain or predict the transfer. The necessary variables for linear regressions are derived from literature research. The models are created both with and without heteroscedasticity adjustment. The results of the models are slightly different, but show a common trend, which originates from the same background formula. Since these characteristics are known in the literature of knowledge transfer, the framework created is well suited for measuring the transfer.