Opened in 2016, the centre allows our psychologists to work with and help autistic people through excellent educational facilities, and by translating theory into practice. Our world-class research team is dedicated to addressing the social and cognitive challenges associated with autism, and making positive changes that facilitate the autistic community’s full involvement in society.
Police interviews are a challenge for autistic people
‘There’s so much going on in this field at the moment,’ says Dr Katie Maras. ‘We’re all doing different but overlapping pockets of work, and I’m concentrating on supporting people with autism in the criminal justice system, with a main focus on interviewing.
‘Autistic people have specific memory difficulties when it comes to recalling specific past episodes in their lives. So while their memory for facts about themselves, or general knowledge, can be amazing, going back in time can be hard and they may struggle.
‘The condition is also associated with problems with social interaction, and a police interview combines both challenges: historical recollection and interaction. Research suggests that these difficulties can be reduced if autistic people receive support in this situation, and that’s what interests me.
‘One of the problems is that we believe autistic people are statistically more likely to be seen as victims or suspects, yet they may not be able to give the forensic detail that’s expected of witnesses in their responses to questions. It’s something that’s investigatively relevant.’
In the 1980s, psychologists devised a method of asking very open questions together with various cognitive mnemonics in interviews, to elicit really detailed reports on complex crimes like murder. This so called ‘Cognitive Interview’ is a technique that robustly enhances the amount of detail most witnesses recall, but it has a detrimental effect on those who are autistic.’
Open questions aren’t appropriate; leading questions aren’t acceptable
Current ‘gold standard’ police interviews depend on open questioning; if officers are seen to have led people in any way, the resulting ‘evidence’ could be thrown out of court. This means they can’t ask questions about a bag, for example, unless the witness mentions one first.
‘The problem,’ says Katie, ‘is that autistic people have huge problems with broad questions. They don’t understand exactly what the other person wants to know, or what’s required of them, and there can also be issues in how they process their memories - this poses a technique issue.
‘We know from lab work that cues can provide more information for autistic people – to the extent that their recall is on a par with non-autistic people – but practically, it’s hard to provide them. When you’re conducting an investigative interview you can’t ask something specific like “Tell me about the fight” unless the witness has already mentioned it, even if you know there was one!
‘In fact, I did my PhD on this topic, because I wanted to examine the suitability of current interview techniques and protocols, and see how theory translated into real life contexts. An experiment in a lab, with static stimuli, is one thing, but it also has to work outside.
‘We set up a simulated event and then interviewed our witnesses independently. There were equal numbers of autistic and non-autistic people, with the latter forming a control group. They were matched, person for person, by age, gender and IQ, so except for one group being autistic they had exactly the same variables.
‘The witnesses started off with ‘mental context reinstatement’, followed by a free recall attempt. This involved asking all the witnesses to close their eyes, then requesting that they travel back in time to re-experience the incident for 10 minutes or so. After that, we asked them to tell us everything they knew. Of course, interviewing is usually a dynamic process but – while it was witness-compatible and based on what each person said – structurally we made it as similar as possible for each person.
‘At that stage, I didn’t come up with any solutions or novel methods of interviewing. However, as a result I’ve spent the last six years letting the police what doesn’t work! And now, after my work at Bath, I’m about to submit a paper on what does!’
Segmenting the crime
‘Before my research, various experimental empirical studies, with basic stimuli like words and pictures, showed that providing more support allows autistic people to recall more accurately,’ explains Katie. ‘The memories are there; it’s just that they struggle to retrieve them. That said, sometimes they can be great witnesses because they’ve remembered a vehicle’s number plate or similar, so obsessions or a detail-focused processing style can be a really good thing!
‘We've recently developed a new technique. We know it’s difficult for autistic individuals to produce an unbound, free flow verbalisation of an entire event, so they inevitably under-perform. So instead, the interviewer will ask an autistic witness to self-segment their recollection at the outset.
‘For example, you might say “What’s the most ‘important’ thing that happened?”, and then, “tell me something else that happened”, writing each answer on a post-it note and sticking it on the wall. You can then revisit each of these parameter-bound topics in turn, asking the witness to freely recall everything they can remember within each segment. After that, they can respond to interviewer prompts with more specific questions in each segment, based on what they’ve already said.
‘By breaking down the event, you allow the witness to focus their search and retrieval strategies in individual segments, reducing the ‘cognitive load’. Displaying the topic boxes on post-it notes also serves as a reminder of the structure of the event, reducing the amount of information they have to hold onto at any one point.
Towards the end, the wall is covered with useful notes that witness can refer to, making it far less daunting for them to describe the event as a whole. We’ve called this interview technique the Witness-Aimed First Account (WAFA)’.
Perceptions and implications
‘One strand of psychology I’m really interested in is perceptions of autistic people. This includes the implications of how other people judge them in the criminal justice system, as well as other contexts such as employment and healthcare.
‘A big problem, for many autistic people, is disclosure of their condition. It’s a major barrier because it makes some of them uncomfortable; they’re not sure when and where to announce it. Why shouldn’t they tell interviewers, I asked, since there are benefits? Was it because of the fear of stigmatisation, or an expectation of poor or unfair treatment?
‘If so, I can offer reassurance. Our research suggests that, if anything, there’s actually a positive bias towards them, over and beyond a non-autistic person in some cases. Whether as witnesses or in the general community, autistic people are perceived as exceptionally honest – after all, lying requires taking into account what’s going on in another person’s mind – and capable of remembering tiny details that most of as can’t recall.
Moreover, there are massive implications in court for witnesses, victims or defendants who decide not to declare their autism, because their behaviour might be mistakenly perceived by jurors as shifty or pedantic. Knowing they’re autistic can shed a different and non-judgemental light on it, and that‘s likely to be beneficial.
This work isn’t just going to help autistic people
‘The point of The Economic and Social Research Council (ESRC) Future Research Leaders scheme is to develop early career researchers, and this is how I’m using my award.
‘So we’ve established that autistic people will struggle to recall things in their entirety, which may be partly related to their inability to work out or read an interviewer’s intentions. But what’s exciting is that, in our initial trial, this new way of questioning increased recall in everyone we tested, autistic or not. In other words, the WAFA interviewing technique seems to work more effectively than standard interview procedures for all types of witnesses.
‘My ESRC project is now broadening this work out into other areas where similar principles apply, such as job interviews and consultations with healthcare professionals. In fact, the latter is particularly relevant because statistically, autistic people are more likely to have medical issues, from gastric problems to epilepsy.
‘It’s a big project to take on, and we’re testing various techniques in the lab. Soon, in the last stages of our work on police interviewing techniques and strategies, Greater Manchester Police will help us by conducting around 240 hours’ of test interviews. The interviews will be trialled in the lab (with mock witnesses to start with), at around the same time as we test our job interview techniques. We’re currently finalising the details and hope that they’ll both, eventually, become standard practice.
‘The academics and practitioners I’ve spoken to are all keen to implement this new way of interviewing witnesses. If you can question autistic people appropriately, clearly you’re going to get better results, and perhaps, in criminal cases, more prosecutions. Also, a carefully structured, productive interview makes investigations more efficient and the process less expensive. Ultimately it requires fewer questions or repeat sessions, which is beneficial for the interviewer as well as the interviewee.
There are so many benefits, because the autistic community possesses a range of valuable but often under-appreciated skills. Better job interviewing will result in more autistic people entering the workforce, where they have a raft of abilities to offer. If businesses can access these skills during that critical recruitment phase, lots of talented people will end up employed, rather than claiming benefits or a jobseeker’s allowance. It’s obviously better – for the individual, the economy and society more broadly – to have good people in work.
‘To sum up, our research shows that more effective interviews will improve just about any situation that combines requirements for drawing on memories and communicating them. There’s more testing to be done, but I really hope we see it rolled out.’
Katie Maras’ work is funded by BA/Leverhulme (grant number SG142540) and the Economic and Social Research Council (grant number ES/N001095/1).