Exploring the Use of Robots for Diary Studies

TL;DR

This paper investigates using a conversational robot as an interactive diary to study in-the-wild human-robot interactions. By developing the Diary Robot system and conducting a seven-day in-home deployment, the authors compare robot-assisted diary entries with traditional text and audio formats, assessing information richness, usability, and compliance. They find that the robot can elicit information comparable to audio diaries and more than text diaries, with high compliance and unique benefits in engagement and depth for certain topics, while noting usability and privacy considerations. The work offers a practical, system-level approach for diary studies in HRI and discusses design implications for deploying embodied data-collection agents in private home settings.

Abstract

As interest in studying in-the-wild human-robot interaction grows, there is a need for methods to collect data over time and in naturalistic or potentially private environments. HRI researchers have increasingly used the diary method for these studies, asking study participants to self-administer a structured data collection instrument, i.e., a diary, over a period of time. Although the diary method offers a unique window into settings that researchers may not have access to, they also lack the interactivity and probing that interview-based methods offer. In this paper, we explore a novel data collection method in which a robot plays the role of an interactive diary. We developed the Diary Robot system and performed in-home deployments for a week to evaluate the feasibility and effectiveness of this approach. Using traditional text-based and audio-based diaries as benchmarks, we found that robots are able to effectively elicit the intended information. We reflect on our findings, and describe scenarios where the utilization of robots in diary studies as a data collection instrument may be especially applicable.

Figures (6)

• Figure 1: In this paper, we explore the use of conversational social robots as instruments for data capture in diary studies, as an alternative to traditional diary methods that use paper, audio, or computer-based data capture.
• Figure 2: Illustrations of the Software Architecture. Once the user initiates free-form chat, they can verbally command the robot to start a diary entry. In the diary entry mode, the system follows a structured flow, going through a set of predetermine questions about bedtime routines. After each predetermined question, possible follow-up questions may be asked, determined by the embedded LLM-agent. All interactions (playing and recording, non-verbal expressions, etc.) are performed through the robot, and all the processing and external API requests are handled by the controller.
• Figure 3: Word count and overall information over the course of the week.
• Figure 4: Distribution of average word count, overall information, and subjective self-disclosures. $**$ denotes $p<0.05$, and $*$ denotes $p<0.1$.
• Figure 5: Total and Unique number of occurrences in diary entries, aggregated at the participant-level. $**$ denotes $p<0.05$, and $*$ denotes $p<0.1$.