Immersive virtual reality technologies have proven multiple experimental benefits for the study of spatial cognition in large-scale environments, such as a high control over variables and safe travels. However, in navigation tasks, the user orients himself thanks to an integrated, unconscious and complex mechanism combining the use of locomotion and cognition. The constraints linked to the use of a motor interface and a virtual environment for navigation raise questions about the possible generalization of the data collected in these digital experiments. Our long-term objective is to evaluate the reliability of virtual reality to assess the user experience in the specific dimension of spatial cognition. However, user experience in virtual reality is complex to evaluate, and navigation tasks are often evaluated with isolated subjective data. In this paper, we describe the design of a comparative study dedicated to spatial cognition and involving a multidisciplinary approach (neurosciences, psychology, and engineering). We present the development of a protocol suited for virtual and real environments, combining the collection of subjective and objective data (behavioral, biomechanical and physiological measurements), then the modelling of a virtual environment dedicated to spatial cognition is described. To finish, the constraints met during an experimental test and the feedback of the participants are presented.