Friday, 24 October 2008

Researching the validity of conferring

As some of the comments in the last thread obviously show, the topic of conferring is a contentious issue at best, at worst it is seen by some people as Police officers lying through their teeth to cover perceived mistakes. Whatever your view is, there is no perfect answer, short of downloading peoples memories into a computer to be analysed, so the best answer has to be one that achieves the highest quality of evidence possible. There has been extensive research in the areas of memory formation and retention which suggests that the current practice of conferring is one of the ways to achieve the best evidence.

In reaction to repeated calls by the IPCC and the MPA, the Met has commissioned Portsmouth University to conduct studies to determine which methods would be best for not only achieving best evidence but also for the welfare of the officers involved. There have been extensive studies already conducted in the US which suggest that the current practice of conferring IS the best method of obtaining accurate evidence, the problem we face is the public simply don't trust that medical and psychological evidence and think that we are colluding when we sit down, debrief to reconstruct the incident and our own individual actions and then write them up.

The following is an article from the Metline magazine published by the Met Police Federation and discusses the outcome of research commissioned by the Met Fed into how errors in recall can occur. Have a read if you want and feel free to discuss, critique or lament et cetera.

The Write Approach - New research sheds much needed light on how the brain processes memory in high stress encounters and calls into serious doubt suggestions that officers should not be able to confer over notes after incidents.

Officers who confer and write a report after a firearms incident are likely to produce a more accurate account of the event than officers who do not confer and are interviewed. That is the conclusion of a research team which carried out an experiment in which volunteer firearms officer were put through a challenging simulation of an incident and were afterwards required to demonstrate what they had remembered of it. The events during the simulation were recorded by three CCTV cameras so that the officers recollections could be checked against what actually happened.

The primary aim of the study was to examine the validity of current theories about how memories are generated and their relevance to Police officers involved in high stress incidents, but it also examined the effectiveness of achieving accurate recall of events from individuals afterwards. The experiment was carried out by the Force Science Institute, which is conducting a number of scientific research programmes for the Met Federation. The researchers realised that the study had to be realistic enough to create genuine stress among the participants, as well as requiring them to make the sort of 'action decisions' that would be vital in a real life encounter. It also had to be arranged in a way that the outcome would be changed by the officers actions and novel enough that the officers would have to read the situation rather than rely on experience.

The answer was to rig the gym at Hendon to resemble the reception area of a hospital ward, complete with waiting patients sitting around. The officers were given a scenario, learning that they were being posted as guards for a man who had been wounded in an armed robbery and was in a locked ward. As they walked through the reception area to their posts, they were confronted with a conflict - which they had not been told about - between the 'brother' of the wounded man and the receptionist (the receptionist having denied the brother access to the ward) of such an intensity that the officers were obliged to intervene. While their attention was taken up by this incident a hostage taker burst through a door beside the reception desk with his left arm around a hostage and a double barrelled shotgun in his right arm. The hostage taker discharged his weapon twice into the floor before pointing it at the officers, who then needed to control or shoot the man. Once he had been disabled, the action ended.

Forty six officers participated in the study, divided into teams of two or three. When it came to recalling the incident some were asked to write reports after conferring with colleagues, some wrote a report but without conferring, some were questioned - after conferring - by a third party while a fourth selection of officers were each interviewed without conferring among themselves.

The researchers, meanwhile, categorised the officers recollections into two types. The first type - called 'internal' - encompassed recollections of the officers own thoughts and behaviour. The second - 'external' - dealt with recollections about the other players and the environment. Each of these types was then further divided into "narrow" (a specific focus of attention) and 'broad' (a wider overview).So, for example, if an officer remembered thinking that he or she had to move to a pillar, the recollection was graded 'internal narrow', while if they thought they were in danger it was an 'internal broad' recollection. A recollection that the shotgun had two barrels was recorded as 'external narrow' while the memory of someone falling to the floor was 'external broad'.

In this way, it was easy to see what sort of things were most likely to attract an officers attention, and which things they were least likely to pay attention to and, consequently, fail to memorise. The different methods of eliciting their memories - written reports, interviews, conferring and non conferring - were also graded for their effectiveness in respect of the number of things officers could recall and the accuracy of those recollections. Overall, the officers recall of 'external' things was more than 400% better than their memories of what they themselves were thinking or doing. And within the 'external' category, their recollection of specifics ('narrow' focus) was almost twice as good as their recollection of the broader picture.

The researchers report comments: "this 'tunnel vision' or 'selective attention' is characteristic of all humans who engage in this type of encounter... this study indicates that this phenomenon also occurs in well trained Police officers." It continues: "the important element of an 'external' focus of attention particularly an 'external' and 'narrow' focus of attention, is that the process of focusing allows the officers to focus on what is important to them at the time but that they can miss other items that may later turn out to be important. For example, if the officers attentional processes at the moment of the shooting were on the alignment of the gun, they did not note anything about the specific movement of the subject, their clothing, the subjects action towards the hostage etc"

The researchers say that this narrow, external focus facilitates great performance but renders officers 'attentionally blind' to anything they are not focused on. Their memories therefore became less accurate, or non-existent, about things upon which they were not focused. The report points out: "They were quite often inaccurate or unobservant about the hostage and almost never noted the behaviours or action of anyone other than the shooter. They almost never noticed the presence or action of the shooters brother and even other officers."

The researchers cite one officer who described how his two colleagues 'froze' and how he fired his first shot before either of them had drawn their weapons, but the video showed them all drawing their guns almost simultaneously, with one - who was just two feet away - firing immediately afterwards. Another officer spent a great deal of time describing how he pushed an individual out of the way, while the video showed him putting his hands as if to push the person but then recoiling back and taking cover. The different methods of getting officers to reconstruct events after the incident - writing reports after conferring, being cognitively interviewed without conferring - had wildly varying error rates.

When it came to recalling those things upon which the participants were 'externally' and 'narrowly' focused upon - that is things to which they were more attentive - the officers who conferred and wrote reports recalled 314 correct details and only made two errors, representing an error rate of 0.14 per officer. The researchers describe the statistic as "amazing". The error rate per officer of the 'external narrow' recollection of the report writing group who did not confer, although still small, was 0.6 per officer. Officers who were interviewed provided, on average, more than twice the amount of information that the report writers did, but their error rate soared. For example, officers who conferred and were interviewed about their 'external narrow' observations recalled 765 items of which 66 were incorrect - an error rate of 4.5 per officer.

Worst for errors were those officers who did not confer and were interviewed, whose error rate in the 'external narrow' category was 6.6 per officer - that is, 47 times greater than the confer/report group.

Commenting on the big discrepancy between report writing and interviews, the researchers comment: "The primary source of the errors dealt with information on the edge of the constables focus of attention. The constables worked hard to provide accurate information but the interviews apparently led them to expand on items that they were less knowledgeable about." The researchers point out that the primary aim of the experiment was to demonstrate the degree to which 'focused attention' occurred in high stress incidents and the manner in which it affected the officers perception afterwards.

The effectiveness of the different methods of eliciting memory recall was only a secondary study and did not meet the criteria of a full study, which would require more officers to participate. But they conclude that, nonetheless, their findings indicate "reports produced less information but are more accurate in their detail" They add: "when constables are going to have to report on an incident, the most accurate reporting of the details is going to be provided by those constables who had a chance to confer. A further conclusion might be that the less the officers provide about the incident, the fewer errors they will make, especially if they report only on the behaviour that they were most specifically focused on"

The article was based on the research report entitled "A study of the presence of perceptual distortions in firearms officers" and raises some very important points which the study by Portsmouth University will go into much greater depth. The issue regarding statements is always going to be the perceived credibility or ability to trust those submitting the evidence, but the ultimate aim is to provide evidence to a court for trial or coroners inquest. There is only so much error in a statement that a jury will tolerate before bringing the whole statements reliability into question. With regard to the accuracy in statements, the problem we face is to decide if we want an extremely accurate statement whose credibility will always be questioned by some people, or if we want less accurate but more detailed Police witness statements that face being deemed unreliable because of the vastly increased number of errors.

As always, it seems we're damned if we do, and we're damned if we don't.


Thursday, 9 October 2008


Given my last couple of posts have been short cathartic rants, this post is a long one, but hopefully I'll explain why.

Following recommendations by the Metropolitan Police Authority after Stockwell, the IPCC recently called for the practice of Police officers being able to confer whilst making statements to stop. This was in direct response to criticisms of Police by the family of Mark Saunders who was fatally wounded by armed officers after he went off his nut and started shooting into other peoples houses and didn't stop when asked. This is the third time the IPCC has called for the practice of conferring to stop after the tragic deaths of Jean Charles de Menezes and Harry Stanley, both of whom were shot by Police officers and were later confirmed to not be armed.

Nick Hardwick of the IPCC said: "The IPCC welcomes the recommendation in the MPA’s scrutiny report that the practice of officers conferring to make their notes following an incident should be discontinued and procedures put in place to demonstrate that the accounts individual officers give are their best and genuinely independent recollections.”

Mr Hardwick added: “We are confident that the investigations we have conducted into fatal police shootings, are rigorous and capable of withstanding public and judicial scrutiny. But we recognise the concern and suspicion this practice sometimes generates amongst bereaved families and many members of the wider public. That suspicion cannot be in the interests of families or the officers concerned.

“The IPCC has a legal duty to secure and maintain public confidence in the police complaints system. As the public body charged with oversight of the Metropolitan Police Service, the MPA’s support for our recommendation confirms the IPCC’s own view that the public do not have confidence in the current procedure in which police witnesses and civilian witnesses to the same incident are treated very differently

“Both the MPA and ourselves recognise the uniquely difficult and dangerous job performed by firearms officers. The IPCC is clear that its investigators do not treat officers as suspects unless there is evidence that an offence has been committed. We recognise that the firearms officers are lawfully carrying weapons and we do not treat them as suspects in a crime unless there is evidence to do so. However, when the state takes a life, we believe that there must be a rigorous investigation and the families and public are entitled to the fullest possible explanation of what occurred and why. This is the approach we have taken in all 14 fatal shootings we have investigated since 2004.

“The current post-incident procedure limits our ability to obtain the best possible evidence from police officers involved in an incident. Each case is different and the importance of the officers' notes will depend on the other evidence we have available.

"The IPCC also recognises that changing the procedures following fatal shootings has far wider implications and may affect the way the police service gathers evidence for criminal investigations. Current guidance reflects the convention that police officer witnesses to an event are permitted to confer before writing their statements. This is a principle in daily police practice. It is not within the IPCC’s power unilaterally to alter policing practices and we recognise that the Police Federation has strong views on the subject.

"While the courts may, in time, come to a definitive ruling on the question of officers’ notes we think ACPO, the Police Federation and the other police organisations need to quickly recognise the current situation is unacceptable. We think it would be possible to develop post incident procedures that provide reassurance to families and the public that best evidence has been obtained and reassurance to officers that they will be protected from unfair treatment for just doing their difficult and dangerous jobs. We seek to work with ACPO and the Police Federation to do that.”

It is a well known fact that Police officers witness statements and other witness statements are obtained in very different ways, as acknowledged by the IPCC. There are a number of extremely obvious reasons for that and a couple of not so obvious reasons, all of which must be taken into consideration before calling for the practice to end.

Firstly, a number of people have said that Police witnesses should be treated no differently from any other witness. As Police officers, we are trained to take statements from witnesses and write our own, and obviously how much practice you get can affect the level of quality of a statement. When an incident occurs we don't get all the witnesses then sit them in a room on their own and ask them to write their own statements, if we did, the vast majority would be completely un-useable in a court and would be missing massive chunks of information as well as having irrelevant information, hearsay and time lines would be all over the place. When we take statements we have to cover rules established under the caselaw of R v Turnbull and follow the ADVOKATE acronym -

Amount of time the suspect was under observation
Distance between the suspect and the witness
Obstructions to view
Known or seen before
Any reason for remembering
Time elapsed between observation and identification
Error or material discrepancy in description

In addition to getting a full account of what happened, we interview the witness and ask probing and open questions to get specific details which can then in turn be expanded upon. A key area to completely breakdown in the recall is the sense of time and the order in which events occur. This can obviously damage the credibility of a witness if everything is later shown to be wrong and ripped to shreds in a court. It's not impossible to imagine a defence solicitor challenging a witness "Quite clearly this event happened before that one........ if you can't even get that bit right....... how reliable is the rest of your evidence?"

I have seen statements from PCSO'S that have been 10-15 lines long which I have then had to retake, expanding them to over 3-4 pages. It wasn't a failing on their part, just a lack of proper statement training, a lack of understanding of the chain of evidence or of identification requirements and a lack of training in criminal law such as offence wordings and points to prove. Considering that PCSO'S are at least regularly involved in Police incidents and occasionally have to give statements, if we had members of the public write their own following incidents they hardly ever deal with, they would for the most part, be worthless.

The complexity of the involvement or the type incident does change how much you can actually recall. Just walking down the street and seeing a shunt and a heated argument can last just as long as someone coming up to you, screaming in your face and demanding your wallet but your recall would be expected to be wildly different between the two. The more traumatic the incident the harder it is to get certain information, as soon as you get a raised heart rate and stress response kicks in the less your brain actually holds onto, especially if you haven't experienced it before. This is called perceptual distortion and is subject to an extensive body of research by psychologists and physiologists, I'll be putting a separate post up about it later on because it's too significant to be just a two or three liner but I've written a bit about it here.

When we write our own statements we very often write them together and get a brew or some food to eat at the same time as it's a good opportunity to get refs breaks. Because our statements are expected to not only present the incident as we saw it, but are basically the backbone of the file, they should contain all of the relevant data for the incident. This would include exact times, descriptions, relevant history, street names, callsigns, authorities, incident numbers etc, most of which we would not have to hand before or during the incident.

As I said earlier, one of the main things to go in memory immediately and soon after a traumatic event is the sense of time and order of events. Whereas we can interview a witness and expand on specific details to make sure the statement is accurate, it's extremely difficult to do this to yourself. Considering how many police officers we usually have on duty, it would be an incredibly enormous and wholly unjustifiable drain on resources to have to have another Police officer who wasn't involved in the incident to obtain our statements for us. There simply aren't enough Police officers to do that.

A decent statement takes an hour or so (at least) if it is a complex incident and you know nothing about it. For a shift of ten police officers each arresting one person every couple of days, most serious incidents happen around the same time (thanks kindly to Murphy's Law) and involve more than one officer, we'd need to increase the number of police officers by at least 3-5 just to take statements from other officers to allow the investigations to be completed expediently. In order to stop the extra officers from being deployed and tucked up with jobs they would have to be specifically for statement taking - so they might as well not be Police officers - but in today's bureaucratic world they would not be employed just to take statements. Heaven forbid they had a couple of days or even a few hours where they didn't actually do anything and just milled about, the fact that they would have to have other requirements in order to even justify their position means that they would run the risk of not being able to take statements immediately.

By sitting down and debriefing the incident and then writing our statements together we can get all of the other bits of information together that we need for the statement, tie up time lines and generally piece together the incident. I've no idea how many times I've been dealing with something and someone has suddenly popped up and then disappeared just as quickly. Gaps in memory and uncertainty about decisions made at incidents are proven triggers for post traumatic stress disorder and eliminating the causes as soon as possible helps to prevent this extremely damaging psychological injury. I wrote about this a bit more here.

The fact that we deal with arrests and violent incidents all the time reduces the influence of perceptual distortion thanks to stress familiarity but it doesn't extinguish it. A bog standard gobby drunk getting floored and cuffed after a short struggle is far from being on a par with someone trying to put a knife in your neck, or having to use extreme or even lethal force. Only by being able to properly debrief every incident are we able to deal with them, the severe pressures on our time such as dealing with calls and prisoners and subsequent admin mean that the only time we can do this practically is when we put our statements together.


Saturday, 4 October 2008

Is she taking the piss?

I'm in nights mode which means at the present stage I'm knackered and my ability to mince words is zero but this, I'm afraid, just HAS to be said. Jacqui Smith on Boris Johnson's 'invitation' for Sir Ian Blair to remove himself from post -

Speaking on BBC's Question Time, Ms Smith said: "There's a process in place that the mayor chose not to respect"

WHAT? you mean like YOU did with our fucking pay negotiations last year and are still doing by completely destroying the fair process that has existed for 30 odd years, simply so you can control it on your terms?!?

She then went on.....

"What is important when you are both choosing and when you're supporting somebody that you're asking to do a job like that is that you keep party politics out of it"

REALLY?? So you're saying that you'll be willing to give the job to someone who is blatantly right wing in their views and who thinks that the Police should be terrorising real criminals and harassing them until they stop committing crime, move, or kill themselves?? Give me a break you fucking hypocrite. The post of Commissioner is everything about politics and everyone knows it. It shouldn't be, but then you bastards have been trying to get the Police under your thumb for a hundred years, exactly how stupid do you think we are?

Home Secretary, as a serving Police officer and one who wants nothing more than harass drug dealers and burglars and their ilk every single day, I for one cannot WAIT until the Tories get in, for no other reason than you'll lose your seat and you'll have to fuck off back to something you actually know, like teaching, which is the only thing you're actually qualified to do.

I've been pissed off about that comment since I read it on the train yesterday afternoon and now I've said my piece I'm going to sleep.