Abstract:

With the advent of commercially available Mixed-Reality(MR)-headsets in recent years MR-assisted learning started to play a vital role in educational research, especially in STEM. Along with these developments it seems viable to further frameworks and structured design processes for MR-based learning environments.

A frequent problem in STEM education is how novices could approach an experimental problem in a more structured way as is commonly done by the more experienced in the field. MR systems are potentially capable to interactively support and tutor novices such that they have a higher chance than usual to solve a given experimental problem in a way that they are not guided too closely. Starting from this idea we suggest an empirically driven problem- and user-centred design process for MR applications in this realm and exemplify it for a specific experiment and problem set containing an non-trivial electric circuit with capacitors and coils.
In the according qualitative study we first carried out a paper-pencil test in order to divide the sample of participants into novices and experts. Afterwards these participants carried out the experiment while wearing an eye-tracking headset. They were given a certain problem to solve, namely to maximize the time between two visible events in the electric circuit by exploring effects of the different components and interchanging them.

The analysis of potentially advantageous features and functions in the planned MR application was then done on the basis of different methods: First, eye-tracking measures have been analysed with regard to different stages of the experimentation process. Second, semi-structured interviews had been conducted with the participants. Additionally, cognitive load with respect to problem solving was assessed and the individual results of the problem solving process were taken into account. Finally, the preliminary design of the MR application was derived from these diverse data.

By evaluating eye-tracking data we derived the thought-related foci of novices and experts on the basis of the eye-mind hypothesis. Based on this assumption we interpreted the experimental process of both groups in comparison. Some objects had been fixated in a much higher intensity by novices than by experts. The conclusion that long fixated objects are those the participants think intensively about matches with the qualitative interviews afterwards where participants were asked to explain their preliminary approaches to solving the tasks. They were also asked what would have potentially helped them to understand the processes in the circuit and to solve the task in the best possible way. We matched their responses with their eye-tracking data and deduced visualizations for the MR application.

In the further process after implementation of the MR application we carried out a usability-study using the thinking-aloud method to improve the MR application. Although only small a sample was considered in this qualitative approach, it resulted in a number of functional and design improvements of the application which is described in detail in this work. At the end we summarize the design process and give an outlook on the main study to follow.

Keywords:

Augmented Reality, Mixed Reality, Eye-Tracking, Learning Environments.