With the rising prevalence of social robots in public spaces, an increasingly heterogeneous audience of people from across different characteristics (such as age or ethnicity) are possible users. In such diverse interactions, there is a risk of algorithmic bias, which results in a discrimination of certain user groups. Human-robot interaction (HRI) research has thus far predominantly focused on homogeneous samples to describe instances and consequences of algorithmic bias, which is prone to leading to evasive findings. We address this gap by conducting thirteen focus group interviews with a total of N = 92 participants and exploring if and how people have already experienced algorithmic bias in their daily lives. Additionally, we contrast the findings from our socially diversified sample with those obtained from a more homogeneous group of people. Our findings uncover various experiences and coping mechanisms regarding algorithmic bias and demonstrate that examining a more diverse sample reveals findings that would otherwise have remained unnoticed.

Transparent communication about how social robots collect and process personal data is essential, especially as their presence in public spaces continues to grow and their applications become more widespread. While previous research has primarily focused on the content and creation of explainable transparency, less attention has been given to which modalities robots should use to convey transparency in the first place. To close this gap, we examined different single modality approaches for communicating transparency and compared them to various combined modalities for transparency explanation, since these have the potential to convey information more efficiently through multiple channels. We conducted a virtual reality (VR) two-part laboratory experiment in which N = 106 participants interacted with a virtual Pepper robot and had to disclose personal data to it. The study design consisted of six conditions: a control group without transparency communication, two multimodal conditions where transparency declarations were presented through multiple channels, and three unimodal conditions where a single channel was used for transparency communication. The results show that the unimodal group was more effective than both the multimodal and control groups in delivering clear and understandable transparency declarations. This suggests that unimodal approaches to transparency may be the preferable option. This study provides insights into transparency declarations in HRI and offers key takeaways on how transparency can be communicated most effectively.

The deployment of social robots in public spaces has received increased interest over the past years. These robots need to process a wide array of personal data to offer services that are tailored to users’ requirements. While much research has been carried out regarding the creation of explainable content, little research has dealt with how data transparency - as a way to address uncertainty and concerns regarding the handling of personal data - is conveyed to users. To examine the impact of different transparency declarations on trust, performance, and robot perception, we conducted a virtual reality (VR) supported laboratory experiment with N = 53 participants who interacted with a robot in a public setting (a library). The interaction between users and robots was accompanied by information on the handling of users’ personal data using three different modalities (via posters, the robot’s tablet, or verbally). The results imply that, while all modalities are understandable and perceived as useful, there is no preference for any modality. Our findings contribute to HRI research by examining different modalities for transparency declarations, in an effort to foster understandable and transparent processing of data.