Police interviewing of autistic witnesses and defendants: What is best practice?

In this article, Dr Laura Crane, a Research Fellow from City University, Dr Lucy Henry, Professor of Speech and Language at City University, Dr Katie Maras, Lecturer from Bath University and Dr Rachel Wilcock, Senior Lecturer from the University of Winchester, discuss how best to interview autistic people by taking a close look at existing research.

Authors: Laura Crane, Lucy Henry, Katie Maras and Rachel Wilcock

Download a PDF version of this article

Police interviewing of witnesses and defendants with autism: What is best practice?

The Criminal Justice System (CJS) is taking autism seriously. Early this year, there was a debate in the House of Lords in which Baroness Uddin expressed dismay regarding the “extraordinary difficulties” people with autism face in obtaining justice.  Justice Minister Lord Faulks QC explained that specific training for police officers at different levels was under development, and was expected to be ready in 2016.  

It remains to be seen what this training will involve, but one of the most important factors to allow any individual to receive appropriate access to justice is being able to provide evidence to the best of their ability at interview.  

So, what can the research literature tell us about how to interview individuals with autism?  In this article, we focus on the (limited) evidence that exists regarding best practice for interviewing witnesses and defendants with autism.

One of the most widely used and accepted forms of interviewing is the Cognitive Interview (CI) – a technique developed by psychologists based on several important memory processes.  The CI is effective in increasing the number of correct details that a witness reports, but, importantly, without also increasing the number of errors made.  Studies have demonstrated the CI’s effectiveness with several different groups of people, including:

The CI comprises a number of instructions including:

  1. The instruction to report absolutely everything – no manner how small, or how trivial it seems

  2. Reinstating the context at the time of the event – the witness is encouraged to think back and visualise the environment in which the event happened (e.g., objects, people, sounds) and is guided to build up imagery for specific details

  3. Building rapport and “transferring control” of the interview to the witness

  4. Asking “witness compatible questions”, for example by structuring questions in the same order as the witness initially recounted the event.

The success of this technique led to the UK’s Home Office (2002) publishing guidelines that recommended the CI to be used with all vulnerable witnesses, including those with autism. However, recent research has suggested that this technique is inappropriate for witnesses with autism.  Comparing the use of a CI against a structured police interview (an interview without the aforementioned techniques), adults with autism actually made more errors with the CI.  

So, are there any alternative interview methods that would be more effective for adults with autism?

One suggestion is a self-administered interview (SAI) technique.  This is a modification of the CI, but it removes the interactive social element of the interview – allowing witnesses to record their memories themselves, guided by specific instructions and questions.  On the face of it, this technique may appear to lend itself well to individuals with autism.  However, empirical research has shown that the SAI is not an effective technique for individuals with autism:

  1. It does not elicit more correct details

  2. It results in adults with autism providing less accurate information.  

However, one specific aspect of the SAI may be a promising method to elicit more correct details from adults with ASD.  This is the ‘sketch plan’ component, in which individuals are asked to draw (rather than verbally recall) the scene of the event.  Further research has used sketching as a modified technique for reinstating the context with children with autism; by asking them to sketch anything they feel would help them remember what they saw. This has been shown to be effective at improving how much information they can remember, without increasing the number of errors.

However, it is important to remember that no two individuals with autism are alike, and what works for one person on the autism spectrum may not work for another.  There is growing recognition within the CJS that personalised support based on an individual’s specific needs is necessary to support witnesses and defendants with autism.  In 2004, a measure designed to provide such support – the Registered Intermediary Scheme - was piloted in six areas in England and Wales. Following this successful pilot, the scheme was rolled out across England and Wales in 2007.

Registered Intermediaries are communication specialists who are available to support vulnerable witnesses and, in Northern Ireland, defendants.  Intermediaries come from a variety of backgrounds (e.g. psychology, speech and language therapy, education, social work, teaching, nursing), and all complete training organised by the UK’s Ministry of Justice.  Importantly, their role is impartial – they do not work for prosecution or defence – and they can assist at all stages of criminal investigation, including during an interview.  For example, they may set out ‘ground rules’ stipulating the prohibition of certain question types that the individual is likely to struggle with, such as tag questions (e.g., “he didn’t take the phone, did he?”).

Initial reviews of the Registered Intermediary scheme are overwhelmingly positive.  Feedback from the pilot scheme highlighted that, as well as facilitating communication and helping individuals to cope with stress, intermediaries appeared to increase access to justice.  At least half of the cases referred to in the pilot would not have reached the trial stage without an intermediary.  

Encouragingly, recent empirical work has also shown that the presence of an intermediary does not hinder jurors’ perceptions of the child, but actually improves perceptions of the interview as a whole.  Although there is no published research assessing whether the assistance of an intermediary improves the quality or accuracy of the information recalled by witnesses, such a study is currently underway and initial results appear positive.

The role of the intermediary is also important given that police interviewers are not expected to be experts on autism.  Indeed, some police forces in the UK opt to train their officers in detecting vulnerability generally, rather than providing training on different disabilities or mental health conditions.  As intermediaries are matched to witnesses (and defendants[1]) on a case by case basis, they can provide expert knowledge on autism to police officers who are tasked with interviewing witnesses (and defendants) on the autism spectrum.

As well as continuing research into the most effective methods for interviewing individuals with autism, it is important to take account of the views of both the autism community and the police.  We need to find out – from those who have personally or professionally experienced the unique challenges autism brings to the criminal justice system:

  • What is feasible

  • What is effective

  • What needs to be done to improve the process for all involved.  

We still have a long way to go to establish ‘best practice’ for interviewing witnesses and defendants with autism, but this is a key first step towards achieving this goal.


Maras, K. L. and Bowler, D. M. (2010). The cognitive interview for eyewitnesses with autism spectrum disorder. Journal of Autism and Developmental Disorders, 40 (11), 1350-1360.

Maras, K.L., Mulcahy, S., Memon, A., Picariello, F., & Bowler, D.M. (2014). Evaluating the Effectiveness of the Self-Administered Interview© for Witnesses with Autism Spectrum Disorder. Applied Cognitive Psychology, 28(5), 693-701.

Mattison, M. L., Dando, C. J. & Ormerod, T. (in press).  Sketching to remember: Episodic free recall task support for child witnesses and victims with autism spectrum disorder. Journal of Autism and Developmental Disorders. 

Plotnikoff, J., & Woolfson, R. (2007). The ‘Go-Between’: evaluation of intermediary pathfinder projects.

Ridley, A.M., Van Rheede, V., & Wilcock, R. (in press). Interviews, intermediaries and interventions: Mock-jurors’, police officers’ and barristers’ perceptions of a child witness interview. Investigative Interviewing Research and Practice.  Research project website: 

[1] Note: at the time of writing this article, the Registered Intermediary scheme in England and Wales (operated by the Ministry of Justice) was only available to support vulnerable witnesses.  In Northern Ireland, legislation has been extended to support vulnerable defendants as well.  The situation in England and Wales may change following a High Court ruling that defendants should also benefit from the assistance of Registered Intermediaries ([2014] EWHC 1944).

Date added: 22 June 2015

Reviewed: 18 June 2018



Thu, 09/07/2015 - 20:01

Another point is, that especially in victim or witness statements: "increased discrimination of false memories in autism spectrum disorder" in other words, autistics have more accurate memory recall than neurotypicals.

Also, autistics are more inclined to be honest: liars have decreased grey matter autistics have increased grey matter


Sat, 18/07/2015 - 11:36

Although there appears to be a typo in the document:

"What started as a routine fact-gathering task may turn into an unnecessary interrogation because an officer, unfamiliar with the behaviours of ASDs may have had their law enforcement instincts rightfully aroused."

...should more correctly say WRONGfully aroused.