Robot-Mediated Interviews with Children : What do potential users think?

Wood, L. J., Lehmann, H., Dautenhahn, K., Robins, B., Rainer, A. and Syrdal, D.S. (2014) Robot-Mediated Interviews with Children : What do potential users think? Society for the Study of Artificial Intelligence and the Simulation of Behaviour.
Copy

When police officers are conducting interviews with children, some of the disclosures can be quite shocking. This can make it difficult for an officer to maintain their composure without subtly indicating their shock to the child, which can in turn impede the information acquisition process. Using a robotic interviewer could eliminate this problem as the behaviours and expressions of the robot can be consciously controlled. To date research investigating the potential of Robot-Mediated Interviews has focused on establishing whether children will respond to robots in an interview scenario and if so how well. The results of these studies indicate that children will talk to a robot in an interview scenario in a similar way to which they talk to a human interviewer. However, in order to test if this approach would work in a real world setting, it is important to establish what the experts (e.g. specialist child interviewers) would require from the system. To determine the needs of the users we conducted a user panel with a group of potential real world users to gather their views of our current system and find out what they would require for the system to be useful to them. The user group we worked with consisted of specialist child protection police officers based in the UK. The findings from this panel suggest that a Robot-Mediated Interviewing system would need to be more flexible than our current system in order to respond to unpredictable situations and paths of investigation. This paper gives an insight into what real world users would need from a Robot-Mediated Interviewing system.

picture_as_pdf

picture_as_pdf
906947.pdf

View Download

Atom BibTeX OpenURL ContextObject in Span OpenURL ContextObject Dublin Core MPEG-21 DIDL EndNote HTML Citation METS MODS RIOXX2 XML Reference Manager Refer ASCII Citation
Export

Downloads