[cs2818]

I couldn’t agree more. To share a little bit of my experience in this area:

I’m a social robotics researcher and our group has spent several years working in this area, looking at the effectiveness, ethics, and protections (or harms) involved in using different emerging technologies for conducting forensic interviews with children. As a result I’ve become very familiar with the different forensic interviewing techniques used with kids.

One thing I can say is historically some horrible approaches have been common, but there’s tremendous effort today to reform this. However, even with these well designed techniques the research shows forensic interviewers fall back to using problematic methods if they aren’t regularly refreshed on proper techniques. Our research is focused on eliminating the biases and leading interviewers are vulnerable to by using software and even using tightly controlled social robotic interviewers to totally control expressions, body language, etc. during these interviews.

The general progression is ask an open ended question, then if that’s unsuccessful offer multiple choice, then yes-no. Our protocols strictly forbid adding information apart from repeating what a child says in open ended responses for clarification. But even that can be tricky, even through a robot. In summary it’s not an easy thing and it seems most police/social services in the US have relatively few resources dedicated to improving it. Here’s hoping we can improve this soon!