The Human-Chatbot Connection with Professor Rheu

In an enlightening exploration into the intricacies of human-chatbot interactions, School of Communication Professor Minjin (MJ) Rheu, has pioneered research that lays bare the psychological underpinnings of our digital conversations. Focusing on the context of social support, particularly during the taxing times of COVID-19, the study offers a deep dive into how college students grappling with academic stress navigate their emotional landscapes through chatbot interactions. 

Rheu's investigation into the psychological dynamics between humans and chatbots, especially in a social support context, is driven by a deep curiosity about our interactions with technology.  

“My interest in human-machine interaction started when I was doing my master’s... I was curious about to what extent machines can emulate human-to-human conversations and their psychological implications,” Rheu said. This curiosity underpins her research, emphasizing the critical need for chatbots to manage expectations through honesty and transparency. Rheu advocates for this approach, highlighting its importance in fostering more empathetic, transparent, and effective human-chatbot interactions.  

Rheu worked alongside her team, which included Wei Peng (Professor, Michigan State University), Yue Dai (Associate Professor of Media and Communication, City University of Hong Kong), and Jingbo Meng (Associate Professor, Ohio State University), 

Rheu's work unravels the nuanced relationship between user expectations and chatbot efficacy, revealing that chatbots capable of delivering contingent, tailored responses can significantly bolster positive psychological outcomes, such as feelings of emotional validation. 

"The gist of it is that participants did experience positive psychological outcomes after talking to a chatbot that responded 'contingently' to users' inputs," Rheu said, highlighting the potential of personalized digital support systems. 

However, the research doesn't shy away from the pitfalls of generic, scripted chatbot responses, which were found to be detrimental, particularly when users were led to overestimate the chatbot’s support capabilities. 

"Expectedly, generic, scripted responses from chatbots had a detrimental effect on generating those outcomes, but, the interesting part is only when the chatbot 'boasted' about its ability to help." She said. This distinction underscores the importance of managing expectations and the power of honesty in the design and deployment of chatbots. 

Rheu's insights not only highlight the delicate balance required in chatbot communication but also set a new benchmark for the development of digital support systems. By recognizing the significance of managing user expectations and prioritizing honesty, her research paves the way for future innovations that could transform the landscape of mental health support.