The Three Stages of Learning to Master a Sensory Augmentation Device: Activation - Acquisition - Integration

Abstract

By augmenting sensory perception through technology, researchers study how humans use and perceive the novel sensory input provided. Nevertheless, little is known about how learning unfolds over time. To this end, 27 participants trained with an augmentation device (feelSpace belt), which provides tactile feedback about the direction of the cardinal north, over the course of six weeks. During the training phase, we tracked participants progress using two different self-report questionnaires containing Likert-scale and open questions. Experts quantified responses to open questions using a previously established category system. As human raters are known to be susceptible to biases, we later reproduced the expert categorization using ChatGPT-4, finding a high congruence between the two classification approaches. The results suggest a three-stage model that best describes the process of acquiring an augmented sense. During the early activation stage, processing the augmented signal requires effort, induces fatigue, and a heightened awareness of the environment. In the knowledge acquisition phase, participants develop a more detailed cognitive spatial representation containing information about objects and places in relation to their current location. The deep integration stage is marked by participants seamlessly integrating the augmented information into existing perceptual processes and automatic use. The results support the supervised use of large language models (LLMs) for the analysis of qualitative data.