Human-robot interaction: Can you trust a robot?

Human-robot interaction: Can you trust a robot?

Posted by Nico Ouwerkerk on Mon 18 Mar. 2013 - 1 minute read

How do you know if someone with whom you do business is telling the truth? When buying something from a stranger, we are trying to determine if he or she can be trusted, either consciously or unconsciously. Non-verbal behavior can often give us cues about the trustworthiness of a person.

In a recent study by DeSteno et al. two experiments were performed to identify non-verbal cues. In the first experiment, groups of two people interacted with one another other using a predefined script. The interactions were recorded on video and The Observer XT was used to code and analyze behaviors that indicate whether or not someone is telling the truth.

The results revealed that it is very difficult to detect trustworthiness using single cues. However, sets of cues in a specific order can determine a level of trustworthiness.

In the second experiment the state-of-the-art robot Nexi from MIT was programmed to mimic specific human behaviors. A group of test participants interacted with the robot while it gave non-verbal cues that are considered indicative of untrustworthiness, and a control group had the same interaction without these cues.

It was found that the group who interacted with NEXI while it gave non-verbal cues (commonly indicative of untrustworthy behaviors) perceived the robot to be less trustworthy when compared with the control group. 

USNavy-Robot MDS

This video shows an initial test of the MDS (Mobile Dexterous Social) Robot designed and built in collaboration with the MIT Media Lab's Personal Robots Group, UMASS Amherst's Laboratory for Perceptual Robotics, Xitome Design, and Meka Robotics.

In short, this study confirms that humans do in fact judge people on non-verbal behavior and that these behaviors can be identified and quantified. The study also concludes that humans respond the same way to non-verbal cues, regardless of whether the source is a human or robot.

So can you trust a robot? If robot sales people are the future, humans are just as likely to be fooled by their trustworthy behavior as by the robot’s human counterparts.

FREE TRIAL: Try The Observer XT yourself!

Request a free trial and see for yourself how easy behavioral research can be!

  • Work faster
  • Reduce costs
  • Get better data


DeSteno, D.; Breazeal, C.; Frank, R.; Pizarro, D.;  Baumann, J.; Dickens, L.; Lee. J. (2012). Detecting the trustworthiness of novel partners in economic exchange. Psychological Science, DOI: 10.1177/0956797612448793.

Featured photo by U.S. Navy photo by John F. Williams [Public domain], via Wikimedia Commons

Don't miss out on the latest blog posts
Share this post
Relevant Blogs

Relational Agents: the use and testing of avatars for HCI

Relational Agents (RAs) are computer constructs designed to interact with humans in a way that promotes the long-term formation of social and emotional bonds.

Human Factors and Ergonomics research in the spotlight

We want to set the spotlight on new developments and publications in Human Factors and Ergonomics research. From maritime simulation centers to the use of video observations in human factors research.

Machine learning and object detection

Humans are incredibly good at recognizing patterns. Now computers can do it as well, and that can be useful. Read more about machine learning and object detection.