Assessing Practitioner Acceptance in Robot-Assisted Autism Therapy: An Anticipated Usability Model (RAAT-AUM)

This study investigates {practitioners}’ perspectives on the {}{anticipated usability} and acceptance of assistive robots in therapy and clinical contexts. It explores their willingness to use these systems when they understand their operations and feel confident in controlling them. Additionally, it examines whether perceived satisfaction is influenced by the alignment between initial expectations and the robots' described capabilities. A robot-assisted autism therapy scenario was presented through illustrations and text, simulating real-world occupational therapy settings. The study assesses the relationships among user comprehension, preferences, confidence levels, and expectations concerning perceived usability, aiming to inform the development of more user-friendly assistive robots. Participants completed a 114-item questionnaire covering preference, self-confidence, expectations, and perceived usability, along with demographic information. Pearson’s correlation analysis revealed nuanced relationships among these constructs. While self-confidence and expectations significantly influenced perceived usability, preferences did not. {}{Understanding the system significantly changed participants’ expectations, whereas preferences and self-confidence remained stable. The results also confirmed the predictive roles of self-confidence and pre-test expectations on anticipated usability.} From these findings, the study introduces the Robot-Assisted Autism Therapy Anticipated Usability Model, which explains how practitioners’ confidence and expectations determine their acceptance of assistive robots, even at the pre-adoption stage. The model extends usability research to the anticipated dimension of user experience and offers practical applications as a diagnostic tool for evaluating acceptance barriers and guiding early design and training programs.