Bartosz Mazurkiewicz
Informed Route Selection for Experimental Design and the Potential of Free Choices for Spatial Knowledge Acquisition

Navigation assistance systems have become widespread, and many people use them daily in familiar or unfamiliar areas. These navigation systems require the user to follow a predefined route, usually the shortest one. Although this approach can be beneficial in certain situations, this circumstance reduces the user to a passive follower, limiting at the same time the interaction with the environment and active decisions. Furthermore, the literature suggests that the usage of navigation assistance systems may adversely affect spatial cognition and orientation, often explained by the forced attention shifts between the supporting navigation system and the environment. Several approaches focusing on enhancing spatial knowledge acquisition have been proposed over the years to mitigate these effects. These approaches try to engage the user in the decision-making process actively. However, they still force users to divide attention or follow a predefined route. In contrast to this work, they do not incorporate the user's environmental spatial abilities allowing for personalized assistance. The work presents two novel concepts tackling these problems by increasing the user's active participation in the decision-making process. The first concept, Free Choice Navigation (FCN), is a non-visual navigation approach focusing on a compromise between free exploration, the number of issued instructions, and a maximum allowed walking distance. Based on estimated probabilities and environmental spatial abilities, this approach also tries to determine whether the user needs assistance at the next junction. The second introduced navigation approach, Beeline Augmented Reality (BeeAR), is permanently displaying the destination through augmented reality data glasses, compensating for the potential errors in users' orientation estimation of the FCN approach. Both approaches focus on decreasing the amount of information provided to the user and involve the user in the decision-making process. Therefore, these approaches are expected to support spatial knowledge acquisition. Both novel concepts were evaluated through an in situ empirical user study using the state-of-the-art turn-by-turn approach as a baseline. One of the most crucial experimental design decisions for such evaluations is selecting the route(s) within an experimental area that must satisfy experiment-specific criteria. As of today, there is little systematic information about the potential influence of routes on study results. Therefore, an agent-based simulation study demonstrates the potential influence of route selection on study results. Furthermore, it will be scrutinized if the sample size impacts this potential influence. The second contribution of this work regarding the importance of route selection in navigation experiments is the development of a route selection framework for wayfinding experiments allowing for systematic and reproducible route selection. Given the abundance of possible routes and their potential impact on study results, this framework should increase ecological validity as the selected routes best satisfy the experiment-specific criteria, which in turn allows for generalizing the findings as much as possible. Besides selecting routes from a potential route population, the framework enables finding similar routes in different geographic areas, which facilitates study replication in other countries. Moreover, a procedure for finding routes leading likely to results in line with most routes in the experimental area is proposed.