Exploratory user study on verbalization of explanations

Shikhar Kumar, Yisrael Parmet, Yael Edan

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This work presents an exploratory user study of verbalization of explanations of a robot collaborating with a human to increase its understandability. Levels of explanation (LOEs) were developed to address two aspects of understandability, namely what and why. Based on these aspects, we defined two parameters, verbosity and justification and developed three different LOEs-high, medium and low. User studies were conducted to evaluate the influence of these different LOEs on performance and user interaction in a collaborative gear assembly task. Trends revealed that verbalization of explanation can increase the robot's understandability. Only the trust measure resulted in a statistically significant difference between medium and low LOE.

Original languageAmerican English
Title of host publication2024 IEEE 4th International Conference on Human-Machine Systems, ICHMS 2024
EditorsMing Hou, Tiago H. Falk, Arash Mohammadi, Antonio Guerrieri, David Kaber
ISBN (Electronic)9798350315790
DOIs
StatePublished - 1 Jan 2024
Event4th IEEE International Conference on Human-Machine Systems, ICHMS 2024 - Hybrid, Toronto, Canada
Duration: 15 May 202417 May 2024

Publication series

Name2024 IEEE 4th International Conference on Human-Machine Systems, ICHMS 2024

Conference

Conference4th IEEE International Conference on Human-Machine Systems, ICHMS 2024
Country/TerritoryCanada
CityHybrid, Toronto
Period15/05/2417/05/24

Keywords

  • Justification and Understandability
  • Levels of explanations
  • Verbosity

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Human-Computer Interaction
  • Control and Optimization

Fingerprint

Dive into the research topics of 'Exploratory user study on verbalization of explanations'. Together they form a unique fingerprint.

Cite this