Theravive Home

A new recently published study looked at robotic mental well-being coaches for the workplace.

“This study investigated the use and deployment of two different robotic forms to deliver positive psychology exercises over four weeks, and how the robotic form impacts on the coachee’s robotic coach perceptions, robotic coach personality, and robotic coach-coachee alliance,” study author Micol Spitale told us. “We hypothesized that the design features, such as the form, may impact the coachees’ perceptions towards the robotic coach.”

The research team deployed two different robots at a workplace in Cambridge, Cambridge Consulstants Inc. All participants were their employees. The researchers relied on their hypotheses on the human-robot interaction literature and psychology, however they didn’t expect to find such an impactful difference in the perception of the robotic forms.

“Previous research has examined the differences between different robotic forms, however, to our knowledge previous work hasn’t investigated two different robots doing the same task,” study author Mina Axelsson told us. “We wanted to see how the use of two different robots would impact people’s perceptions in the long term.”

The study authors explained that mental health has been worsening worldwide, especially after the impact of COVID-19. Robots could help maintain mental well-being, and have a preventive role.

“We decided to explore more how robotic coaches can promote mental well-being in the workplace because we envision that robots can be used as tools to maintain people's mental health, and can provide another mental health service to meet the high demand,” Spitale told us.

To test their theory, the researchers ran a between-subject study to evaluate how the coachees’ perceptions differ across different robotic coaches form groups. During the study, participants came to interact with the robotic coach once every week, for four weeks. Each session was ten minutes. The robot did different positive psychology exercises with people - savouring, gratitude, accomplishments, and optimism about the future.

The team found that the toy-like robot (i.e., Misty II) was perceived more positively than the human-like robot (i.e., QT robot). The researchers hypothesized that this is due to the form function attribution bias: humans tend to attribute capabilities and functionalities of an object based on its form. In their case, coachees attributed to the QT robot (the more humanoid like robotic coach) more sophisticated capabilities with respect to the ones it was equipped with, and this in turn resulted in a huge mismatch between their expectations and the reality (i.e., real capabilities of the robot).

“We also found that people perceived the two robots’ behaviours to be different, even though they were the same,” Axelsson told us. “Participants perceived the toy-like robot Misty to have a more ‘caring’, ‘warm', and ‘understanding’ personality. QT on the other hand was described in more neutral terms, and as ‘not really having a personality’.”

Though the researchers thought the results were supported by the literature, what they found interesting was that people perceived differences in the robots behaviours, even if the only difference is in the forms.

“These results can inform the design and deployment of robotic well-being coach and they contribute to the vision of taking robotic coaches into the real world,” Spittale told us.

“Since the form of a robot impacts our perceptions of its other capabilities, this helps us further understand how complex the process of designing a robot really is!” Axelsoon told us.