Enhanced Prediction Models for Predicting Spatial Visualization (VZ) in Address Verification Task

In the field of Human Computer Interaction and Psychology, it is accepted that spatial visualization (VZ) is one ability that can indicate individual’s performance on computer applications. Since users with different levels of VZ seem to prefer different types of user interfaces (UI), knowing a user’s level of VZ provides a great opportunity for application developers to design software with higher satisfaction and usability. In this paper, we proposed three models to predict a participant’s level of VZ based on the participant’s actions (taps) on the tablet screen while doing an address verification task in the neighbor- hood using the tablet. After applying the proposed prediction models with data of thirty participants, they yielded an optimal accuracy of 93.33%.