Impact of Robot Actions on Social Signals and Reaction Times in HRI Error Situations
Open accessAbstract
Human-robot interaction experiments featuring error situations are often excluded from analysis. We argue that a lot of value lies hidden in this discarded data. We analyzed a corpus of 201 videos that show error situations in human-robot interaction experiments. The aim of our analysis was to research (a) if and which social signals the experiment participants show in reaction to error situations, (b) how long it takes the participants to react in the error situations, and (c) whether different robot actions elicit different social signals. We found that participants showed social signals in 49.3% of error situations, more during social norm violations and less during technical failures. Task-related actions by the robot elicited less social signals by the participants, while participants showed more social signals when the robot did not react. Finally, the participants had an overall reaction time of 1.64 seconds before they showed a social signal in response to a robot action. The reaction times are specifically long (4.39 seconds) during task-related actions that go wrong during execution.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Giuliani et al. [115,116] found that participants showed more social signals when robots violated social norms compared to when they exhibited technical failures.
“In their further investigations [115,116], they also found that participants showed more social signals when robots violated social norms compared to when they exhibited technical failures.”
From A Taxonomy of Social Errors in Human-Robot Interaction (2021) · cites this paper as [115]
Cites
Updating freeze: Aligning animal and human research
Hagenaars, M. A., Oitzl, M., & Roelofs, K. (2014) · Neuroscience & Biobehavioral Reviews · doi:10.1016/j.neubiorev.2014.07.021 · Cited by 260 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Honig, S., & Oron-Gilad, T. (2018) · Frontiers in Psychology · doi:10.3389/fpsyg.2018.00861 · Cited by 338 · see citing works
To err is robot: how humans assess and act toward an erroneous social robot
Mirnig, N., Stollnberger, G., Miksch, M., Stadler, S., Giuliani, M., & Tscheligi, M. (2017) · Frontiers in Robotics and AI · doi:10.3389/frobt.2017.00021 · Cited by 244 · see citing works
Self-citationA Taxonomy of Social Errors in Human-Robot Interaction
Tian, L., & Oviatt, S. (2021) · ACM Transactions on Human-Robot Interaction · doi:10.1145/3439720 · Cited by 133 · see citing works
Kox, E. S., Siegling, L. B., & Kerstholt, J. H. (2022) · International Journal of Social Robotics · doi:10.1007/s12369-022-00871-4 · Cited by 22 · see citing works
- Authors
- Mirnig, N., Giuliani, M., Stollnberger, G., Stadler, S., Buchner, R., & Tscheligi, M.
- Affiliations
- University of Salzburg (AT)
- Corresponding author
- Mirnig, Nicole
- Venue
- Lecture Notes in Computer Science
- Publisher
- Springer International Publishing
- Published
- 2015
- Pages
- 461-471
- ISSN
- 0302-9743 (ISSN-L)
- 1611-3349
- OpenAlex
- W2277838876
- Cited by
- 26
- References
- 1
- Citations by year
- 2026: 3 · 2025: 3 · 2024: 4 · 2023: 2 · 2022: 3 · 2021: 3 · 2020: 3 · 2019: 2 · 2018: 1 · 2017: 1 · 2016: 1
- FWCI
- 1.76
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Lecture notes in computer science (publishedVersion, cc-by-nc)
- Topics
- Social Sciences › Psychology › Social Psychology › Social Robot Interaction and HRI
- Physical Sciences › Computer Science › Artificial Intelligence › Speech and dialogue systems
- Social Sciences › Psychology › Experimental and Cognitive Psychology › Emotion and Mood Recognition
- Keywords
- human-robot interaction, social signals, reaction time, robot action, social norms violation, technical failure, nonverbal communication
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review