Abstract
Maze tasks, originally developed in animal research, have become a popular method for studying human cognition, particularly with the advent of virtual reality. However, these experiments frequently rely on simplified environments and tasks, which may not accurately reflect the complexity of real-world situations. Our pilot study aims to transfer a multi-alternative maze with a complex task structure, previously demonstrated to be useful in studying animal cognition, to studying human spatial cognition. The challenges to be resolved at this stage included developing a virtual maze and selecting an appropriate instruction that will elicit processes similar to those observed in animal models. A virtual maze was developed, and two types of instructions were provided to the participants: (1) to collect coins; (2) to interact with the maze in order to draw its structure after the game. The results indicate that a more structured instruction with a clear attainable goal ("collect") prompted more in-depth exploration and engagement with the key elements of the maze, eliciting processes similar to those of animals. While the maze demonstrates promise as a tool for comparative studies, it also has the potential to uncover different aspects of human cognition.