Conformity in Human–Robot Interaction

For decades, research has explored if and how robots can persuade or influence human decision-making and behaviour. Here, an important facet is that of conformity, which is a social phenomenon that occurs when humans adapt their beliefs or behaviors to match those of other people or entities, typically within a particular social group. As a wider uptake of social robots is becoming increasingly possible in settings such as education and healthcare, this project seeks to understand the mechanism by which such robots can influence humans (particularly children) and the ethical implications thereof.

Background

The study of conformity in human groups dates back to the 1950s when Salomon Ash carried out his famous conformity experiments, which showed that people knowingly provide incorrect answers to an unambiguous task when under the influence of other people. Similar findings have been replicated when replacing human confederates with robots. Hence, humans can be susceptible to social influence whether it be from people or robots to some extent. Yet, little is known when it comes to how, in what contexts, and why these phenomena occur.

As social robots are slowly but surely increasing in sophistication, it is possible that these will constitute everyday elements in future schools, hospitals and workplaces. Such a reality may lead to a range of educational opportunities, but will inevitably also raise ethical issues around robots’ social influence on people, in particular when it comes to vulnerable populations such as children.

Purpose

The aim of this project is to explore when, how and why (or why not) children and adults conform to social robots. Questions relating to different robot embodiments and behaviors will be investigated, as will the impact of individual differences between people such as age and previous technology experience.

Planned activities

The project will:

• conduct a scoping review of the literature on conformity in HRI,
• develop an experimental setup consisting of a variety of robot embodiments, including, e.g., the NAO, Furhat, Epi, and Misty II, and based on previous research,
• carry out experimental and qualitative studies with children.