Equipping social robots with fitting personality characteristics has been shown to increase user enjoyment (Lee et al., 2006) and users’ performance during human-robot interactions (Tapus et al., 2008). However, since many of the application contexts of social robots are public (education facilities, trade fairs, museums, etc.) (Mubin et al., 2018), more than one robot will likely be deployed simultaneously to be available for as many people and tasks as possible. This poses the question of whether the personalities within such robot groups should be uniform or diverse. On the one hand, a uniform personality design could create the best possible interactions if certain favored perceptions associated with personality traits are to be generated on the user side. (Andriella et al., 2022). On the other hand, this uniformity could reduce the social attraction of a robot group and users’ enjoyment of the interaction, as the robots are more likely to be perceived as mere clones of each other and consequently as less unique and having less agency (van der Woerdt & Haselager, 2019; Wiese et al., 2017), which could result in a decreased motivation to use the robots again.

Since these potential effects of uniform and diverse personalities are highly relevant but have not been investigated, they were tested in a between-subjects experiment using virtual reality. Participants interacted with three robots of the same model (Pepper) equipped either with different personality characteristics (extrovert, ambivert, and introvert) or the same personality characteristics (randomized either only extrovert, ambivert, or introvert). Pre-tested personality characteristics were manipulated by altering the robots’ gestures, wording, volume, and movement speed. By measuring participants’ perceptions of the robots’ agency and uniqueness, their effects on social attraction, enjoyment of the interaction, and reuse intentions were tested to deduce recommendations for effectively managing personality characteristics in robot groups.

The increasing integration of robots into society leads to human-robot interaction in groups. In multi-user scenarios, robots must allocate resources, which can lead to disadvantages due to algorithmic biases (Denis Newman-Griffis et al., 2023). Discriminatory robot behavior often goes unnoticed (Hitron et al., 2023), as their actions appear objective and fair. However, it can negatively impact perceptions of the robot’s likability and intelligence (Büttner et al., 2023) and have emotional consequences for users (Spisak & Indurkhya, 2023). Moreover, first empirical evidence hint, that discrimination by robot also effects the interaction partners’ perception of the discriminated user regarding competence (Straßmann et al., 2024). This effect could lead to problems in teams, particularly in work situations where there is pressure to perform and time pressure. Time constraints can limit group interactions, negatively affect collaboration (Karau & Kelly, 1992), and increase competitive behavior (Saraiva & Iglesias, 2015). Thus, time could foster the negative consequences of discriminatory robot behavior.

Using a mixed reality experiment with a 2 (time pressure vs. no time pressure) x 2 (disadvantaged vs. non-disadvantaged) between-subjects design, this work investigates how time pressure influences psychological consequences of discriminatory robot behavior in human-human-robot interaction. Two same-sex participants sorted books together in the University Library with the virtual robot Pepper. The dyads either have been told that the time was being measured during the task (time pressure) or received no information on time constraints (control group). In each dyad the robot made more technical errors when interaction with one person (90%) during the book scanning process than in the interaction with the other (20%). Questionnaires measured perceptions of the robot (Godspeed), warmth and competence of the interaction partner, attribution of blame and success, and emotional effects (NTS, PANAS). Open-ended questions addressed the interaction partner’s perception and feelings of time pressure.

Social robots are increasingly investigated for their potential in mental healthcare (Guemghar et al., 2022), to counter personnel shortages. This study aims to investigate them as therapeutic assistants for adolescents with mood disorders, such as depression, awaiting treatment. A recent review indicates digital communication technology’s potential for this purpose but shows that only two conversational agents have been tested in randomised controlled trials (RCTs), with no robots evaluated (Dietvorst et al., 2024). Research also shows that a physical robot supporting stress-reducing disclosure interactions for adolescents (14-21 years) led to greater stress reduction than digital or VR representations of a robot (Björling et al., 2021), emphasising the potential importance of embodiment and physical presence. To prepare RCTs, we examine the following research questions:

RQ1: What needs of adolescents with mood disorders can be supported by social robots?

RQ2: What features or specific robot characteristics are most suitable?

We will apply a two-stage qualitative approach. First, we will interview mental healthcare professionals to assess a) the target group’s current challenges; b) existing treatment approaches; c) views on potential support by social robots; and, d) integrating social robots into the care process. Insights will inform the design of robot interaction scenarios for the second stage.

Second, we will conduct focus groups with the target group (three age groups: 12-15, 16-17, 18-23), examining their support needs and current support/coping mechanisms. Participants also engage in various scenarios with multiple robots, reflecting on their preferences and the features of these robots and their views on potential robot support.

Furthermore, socio-demographic information (e.g., age, gender, robot experience) will be collected. Audio recordings will be transcribed anonymously and analysed using reflexive thematic analysis (Braun et al., 2023) in Atlas.ti 25. Adolescents will also be asked to record their first impression on paper right after robot interaction.