1.2 Reading minds and behaviour
The research you’ve explored in this course has been based on measuring eyewitness behaviour – such as identification accuracy and the amount of information recalled in an interview. However, it is possible to study the brain more directly by using imaging techniques.
In the next video, Hayley Ness considers what insights might be gained into the thought processes of an eyewitness by measuring the activity in their brain.
Download this video clip.Video player: ou_futurelearn_psychology_vid_1031.mp4
Skip transcript
Transcript
DR HAYLEY NESS
When a person has witnessed a crime, they can often be asked to work with a police operator to construct a facial likeness of the person and we call them facial composites and the most common term for those is an E-fit. There are a number of issues with constructing a social composite, the main one is that it is an incredibly difficult task. So you know, if I ask you to remember the eye shape, brow shape, size, distance apart, you know of the eyes and eyebrows of someone that you know really well, it would be incredibly difficult to describe that but that's exactly what we ask an eyewitness to do and they've only seen the person probably for about thirty seconds or a minute. So it's almost an impossible task.
So the original systems like Photofit and Identikit they were developed on the premise that we can perceive, store and remember individual features. Lots of research in the 1970s showed that there were numerous problems with these systems. Not only in that process but also in recognising an image because the aim of a composite is to show it in the media in the hope that somebody who is familiar with the person will recognise them and come forward. But Photofit and Identikits had lines in them where individual features came together, they had differences in contrast and texture and that all interfered with the ability to later recognise the image.
So E-FIT and Pro-fit were developed and they were seen as a huge improvement on the older systems, so they still ask a witness to process individual features but they're shown in the context of a whole face. So they are an improvement and you can construct recognisable likenesses with them but there are still issues. So as psychological research has developed and face processing and memory research has developed then that's been used to further improve the systems. The problems I've been witnessing have been slightly different I haven't been so interested in development in technology.
In fact I've probably been one of the people or maybe the only person who said hang on a minute can we stop developing technology because we don't actually really understand what a witness is doing when they construct an image. Once we do understand that, then we can maybe start to develop even better composite systems.
My research, its starting point was really how do we find a method that will allow us to find out what a witness is actually doing when they construct a facial composite. So I looked to psychological techniques and in particular, neuro-psychological techniques and I decided to use a method called ERPs (Event Related Potentials) and essentially this just involves asking a participant to wear a cap, its got electrodes on. It's completely painless, there's 64 electrodes on the cap that we used and then we sat them down in front of an automated composite system. So they were shown a Pro-fit face, the features change automatically and they were asked to decide whether the features they were presented with were similar to the target person they'd seen earlier or not and then we measured the EEG when they were making that decision. And this tool allowed us to be able to measure the perceptual and attentional, and memory processes that were occurring.
It gives a lot of data so we use psychological research to hone down what we were looking at so looked to see if there were key face processing markers there. So were people processing the image in the same way that they would process an everyday image of a face. Were they using memory processes, were they attending to the whole face rather than just to the feature and were there differences across individuals. And one of the striking things we found was that there were quite large differences across individuals and particularly at the early perceptual stages and we expected to find differences but we expected that people's memory would differ but what we found was that they differed at the very early stages, they were looking at features in a different way. It wasn't that they were remembering them in a different way. What it means is that some people may favour those older systems, they may produce better images using E-FIT. Whereas, other people may produce better images using the newer more holistic systems. And it may not just be a case that everyone will create a better composite using the newer systems.
Interactive feature not available in single page view (see it in standard view).