Implicit robot bias: differential amygdala activation to subthreshold and suprathreshold robots
Poster No:
814
Submission Type:
Abstract Submission
Authors:
Zhengde Wei1, Ying Chen1, Xiaochu Zhang1
Institutions:
1University of Science and Technology of China, Hefei, Anhui
First Author:
Co-Author(s):
Introduction:
As robots increasingly become part of our daily lives, understanding human-robot interactions is imperative [1-2]. An increasing body of studies have addressed this issue of evaluations of robots with self-reports; yet evidence for evaluations of robots has been inconsistent, ranging from enthusiasm to anxiety and fear of robots [3-5]. A potentially more direct and useful approach for understanding the fundamental processes underlying controlled and automatic evaluations of robots is to probe brain response to different perception levels of robot-related stimuli.
Methods:
In Experiment 1, we recruited 66 participants and asked them to complete an attitude toward robots questionnaire and an implicit association test (IAT; Figure 1a) with image stimuli of humanoid robots and Caucasian individuals. In Experiment 2, we scanned 59 participants with functional magnetic resonance imaging (fMRI) while they completed a backward masking task, in which they viewed suprathreshold (duration: 200 ms) and subthreshold (duration: 17 ms) presentations of humanoid robot and Caucasian stimuli. In Experiment 3, Out of the 59 participants in fMRI experiment above, 23 participants proceeded to complete a positive evaluative conditioning task. Afterwards, they repeated the backward masking task while undergoing fMRI scanning.
Results:
Self-reports and behavioral measures demonstrated a discrepancy between explicit positive and implicit negative attitudes toward robots (Figure 1b-c). Consistent with negative implicit attitudes, fMRI during a backward masking task showed heightened responses in the left amygdala to subthreshold (implicit) as opposed to suprathreshold (explicit) robot images (Figure 2a-c). Importantly, by pairing robot images with positive stimuli, we were able to reduce negative implicit attitudes, which corresponded with attenuated amygdala response to subthreshold robot images (Figure 2d-f).
Conclusions:
Our studies offer compelling evidence for distinct patterns of amygdala activation between controlled and automatic processing of robots. This sheds light on how people evaluate and interact with robots that are increasingly entering our social environment. The future of social robots is undeniably exciting, and insights from neuropsychology research will guide the future direction of robot development and bring us closer to interacting with social robots.
Emotion, Motivation and Social Neuroscience:
Social Interaction 1
Social Neuroscience Other 2
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI)
Novel Imaging Acquisition Methods:
BOLD fMRI
Keywords:
Other - robots; social evaluation; amygdala
1|2Indicates the priority used for review
Provide references using author date format
2. Bossi, F. et al. The human brain reveals resting state activity patterns that are predictive of biases in attitudes toward robots. Sci Robot 5, eabb6652 (2020).
3. Naneva, S., Sarda Gou, M., Webb, T. L. & Prescott, T. J. A Systematic Review of Attitudes, Anxiety, Acceptance, and Trust Towards Social Robots. Int J of Soc Robotics 12, 1179–1201 (2020).
4. Latikka, R., Savela, N., Koivula, A. & Oksanen, A. Attitudes Toward Robots as Equipment and Coworkers and the Impact of Robot Autonomy Level. Int J of Soc Robotics 13, 1747–1759 (2021).
5. Graaf, M. M. A. de, Allouch, S. B. & Lutfi, S. What are people’s associations of domestic robots?: Comparing implicit and explicit measures. in (2016). doi:10.1109/ROMAN.2016.7745242.