“You Scare Me”: The Effects of Humanoid Robot Appearance, Emotion, and Interaction Skills on Uncanny Valley Phenomenon
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsDear Authors,
1. Use all 5 keywords from the following list:
https://www.ieee.org/content/dam/ieee-org/ieee/web/org/pubs/ieee-taxonomy.pdf
It does not add novelty to the subject compared with other published materials.
2. Figures should have left alignment.
3. The conclusions are quite consistent, they have some evidence and arguments presented they address the main question posed. But the conclusions can still be improved. Make a separate "Discussions" section before "Conclusion".
4. Make a "Notations and Abbreviations" section before "References".
5. The references are appropriate. But some references are quite old, e.g. ref. [43].
6. The biggest lack of paper is that has no scientific soundness, there are not used any formulas. Why are there no formulas, mathematics and equations in the scientific paper? More formulas, equations and mathematics should be added to the paper to increase scientific quality of the paper. The paper looks like a report, not a scientific paper.
Best Regards
Author Response
Dear Reviewer1, thank you for your feedback, please find below our responses:
Comment 1: 1. Use all 5 keywords from the following list: https://www.ieee.org/content/dam/ieee-org/ieee/web/org/pubs/ieee-taxonomy.pdf . It does not add novelty to the subject compared with other published materials.]
Response 1: I have replaced all keywords except Uncanny valley from the list as I think this is the selling point of our work.
Comment 2: Figures should have left alignment.
Response 2: We have adapted all figures to be left aligned using latex code: \begin{figure} \raggedleft
\includegraphics[width=.6\textwidth,left]{example-image}
\caption{left aligned image}
\end{figure}
Please check if this is what you expected, currently some figures such as Figure 1 looks strange due to this alignment.
Comment 3: The conclusions are quite consistent, they have some evidence and arguments presented they address the main question posed. But the conclusions can still be improved. Make a separate "Discussions" section before "Conclusion".
Response 3: We would appreciate if you could please provide constructive feedback how conclusion can be improved.
Comment 4: Make a "Notations and Abbreviations" section before "References".
Response 4: We have added this now.
Comment 5: The references are appropriate. But some references are quite old, e.g. ref. [43].
Response 5: Ekman was the pioneer in Facial Emotion Coding therefore we think it is necessary to be included being a fundamental work in emotions and human response.
Comment 6: The biggest lack of paper is that has no scientific soundness, there are not used any formulas. Why are there no formulas, mathematics and equations in the scientific paper? More formulas, equations and mathematics should be added to the paper to increase scientific quality of the paper. The paper looks like a report, not a scientific paper.
Response 6: This research article is presented as a User Studies article, one including laboratory, field studies, and online surveys with real-world robotic systems. Unfortunately this means that we have no formulas to state our findings as the findings are what users attitude towards such systems.
We hope this answers your questions.
Best regards,
Ashita Ashok
Prof. Karsten Berns
Reviewer 2 Report
Comments and Suggestions for AuthorsThis is very important issue. My only suggestions is to include psychologists in future research team. I strongly believe this can improve research quality.
Author Response
Dear Reviewer2, thank you for your feedback, please find below our responses:
Comment 1: This is very important issue. My only suggestions is to include psychologists in future research team. I strongly believe this can improve research quality.
Response 1: As mentioned in our article, we have approval from Ethics Committee of the Department of Social Sciences, University of Kaiserslautern (RPTU Kaiserslautern) which we received due to regular collaboration with Cognitive Psychologists like Prof. Daniela Czernochowski and Dr. Ann-Kathrin Beck.
We hope this answers your questions.
Best regards,
Ashita Ashok
Prof. Karsten Berns
Reviewer 3 Report
Comments and Suggestions for AuthorsThe subject paper is very interesting and the paper is quite clear. The comments are as follows:
1) In the abstract, authors must underline the novelty of their contribution compared to the research activity of others.
2) All figures and tables should be added after their citation in the full text.
3) Subsections 2.1, 3.1, 4.1 and 4.2 need to be better introduced and explained. They consist of only one question!
Author Response
Dear Reviewer3, thank you for your feedback, please find below our responses:
Comment 1: In the abstract, authors must underline the novelty of their contribution compared to the research activity of others.
Response 1: We have modified the last sentence of the abstract from “This research contributes to understanding the complex dynamics at play in the uncanny valley phenomenon and the crucial role of effective humanoid design and behavior in enhancing human-robot interaction (HRI)” to “This research advances our understanding of the uncanny valley phenomenon and the role of humanoid design in enhancing Human Robot Interaction, marking the first direct comparison between the most advanced, human-like research robots.”
Comment 2: All figures and tables should be added after their citation in the full text.
Response 2: We have modified this please check the new version attached below
Comment 3: Subsections 2.1, 3.1, 4.1 and 4.2 need to be better introduced and explained. They consist of only one question!
Response 3: The goal with introducing a research question in user studies is important because it helps focus the study and shows readers what the research is trying to solve or understand in the following sections thereby we clearly stated it separately.
We hope this answers your questions.
Best regards,
Ashita Ashok
Prof. Karsten Berns
Reviewer 4 Report
Comments and Suggestions for AuthorsThe paper investigates the impact of humanoid robot appearance, emotional expression, and interaction skills on the Uncanny Valley phenomenon using the Social Humanoid Robot (SHR) Ameca. The research was conducted in two key studies: one in a university hallway to assess student interactions and expectations and another comparing human-likeness across three robots—ROMAN, ROBIN, and EMAH. The results highlighted the complexities of human-robot interaction (HRI) and the importance of robot design in enhancing human acceptance.
Strengths of this study include the following:
1. Comprehensive Study Design: The use of both field and online studies provides a robust methodology for capturing a wide range of human-robot interactions.
2. Detailed Analysis: The paper offers an in-depth analysis of human responses to robot appearance and behavior, contributing valuable insights into the Uncanny Valley phenomenon.
3. Relevance to HRI: The research is highly relevant to the field of social robotics, particularly in understanding how robot design influences human acceptance.
4. Diverse Participant Pool: The study involved participants from various cultural and linguistic backgrounds, enhancing the generalizability of the findings.
Several limitations of this study include the following:
1. Limited Sample Size: The relatively small number of participants (24 in the field study and 25 in the online study) may limit the generalizability of the findings.
2. Technical Limitations: The study acknowledges malfunctions in the ROMAN and ROBIN robots, which may have affected the participants’ perceptions and the overall results.
3. Lack of Longitudinal Data: The study provides a snapshot of human-robot interactions, but it does not explore how these perceptions might change over time with increased exposure to robots.
4. Privacy Concerns: Participants expressed significant concerns about data privacy, which could impact the acceptance of robots in social settings.
Recommendations to Improve the quality of the paper are as follows:
1. Increase Sample Size: Future studies should involve a larger and more diverse sample to enhance the reliability and validity of the findings.
2. Address Technical Issues: Ensuring that all robots are fully functional during experiments will help produce more accurate results.
3. Incorporate Longitudinal Studies: Conducting follow-up studies over a longer period could provide insights into how perceptions of human-likeness and the Uncanny Valley effect evolve over time.
4. Enhance Privacy Measures: Addressing data privacy concerns by implementing strict data protection protocols and clearly communicating these to participants could improve acceptance of humanoid robots.
Author Response
Dear Reviewer4, thank you for your valuable feedback, please find below our responses:
Comment 1: Limited Sample Size: The relatively small number of participants (24 in the field study and 25 in the online study) may limit the generalizability of the findings.
Response 1: This user study research on uncanny valley effects of humanoid robot presents two separate works conducted with Ameca robot. The first study contains two sessions, a field study with 24 participants and an preliminary online survey with 25 students. Based on the results from the first study we developed a second main online study with 70 participants (3 data points had to be deleted due to not matching ethics approval of age group 18-35). Therefore we can state that our results are a good starting point for attitude of university students towards human-like robots.
Comment 2: Technical Limitations: The study acknowledges malfunctions in the ROMAN and ROBIN robots, which may have affected the participants’ perceptions and the overall results.
Response 2: We think there is a misunderstanding, the second study conducted via online platform did not mention or highlight existing issues with ROMAN and ROBIN robots. The only possible effects might be that the video quality of ROMAN was low as it was one of the older robots from our research group (dis-assembled due to parts malfunctioning). The video links are provided in the manuscript to help understand our point better. But I would like to highlight that we made the best of what we had for a comparison across three robots even maintaining male voice of robot with male interaction partner. This can be seen better in Section 3.2 where we mention “Unfortunately, compared to EMAH due to robot malfunctions old videos of ROMAN, and ROBIN were used, thereby adding another variable - the content and scenario of interaction.”
Comment 3: Lack of Longitudinal Data: The study provides a snapshot of human-robot interactions, but it does not explore how these perceptions might change over time with increased exposure to robots.
Response 3: We agree this is the case with the current work measuring first impression of human-like robots. We must clarify that we have previously conducted several individual studies with Robin, and Ameca, and one longitudinal study with Ameca (4 weeks) which was accepted as abstract, with full submission due December 2024 in "Designing and Generating Effective Behaviors for Social Robots" in Frontiers in Robotics and AI journal. We will definitely conduct more longitudinal studies and hopefully measure decrease in uncanny effect due to reduction in novelty effect.
Comment 4: Privacy Concerns: Participants expressed significant concerns about data privacy, which could impact the acceptance of robots in social settings.
Response 4: Due to the ethical rules that we must follow for safe HRI we have never recorded any form of re-identification human data like capturing voice or images. Only Likert responses to questionnaires in social robotics. As we aimed to gather first impression of human-like robot we did not highlight this part but in other individual in person lab studies we inform participants that no data is recorded except their questionnaire responses which are very generic to surveys.
We hope this answers your questions.
Best regards,
Ashita Ashok
Prof. Karsten Berns
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsDear Authors,
1. some figures still have no left alignment, but maybe this can be also corrected with the editorial team.
2. I would still prefer some formulas and plots, to increase scientific soundness, you can't convince me that in a humanoid robot, no formulas were used. Some movement kinematics at least.
Best Regards
Reviewer 4 Report
Comments and Suggestions for AuthorsThe authors' responses met my expectations, and I am accepting the paper in its current form.
