The Art of Communication

Something else I have been thinking about over the Christmas break is communication.

Ollie criticised my pseudo-3D graphs showing the differences in time profiles of people of different ages who were victims of street robberies. He kindly pointed me in the direction of ""Tufte's seminal work "The Visual Display of Quantitative Information"". Unfortunately I have not been able to find an e-book version of this work on the Internet so I have yet to read it but I have found plenty of references to it and even extracts from it. It looks like it is a book I should have put on my Christmas present list!

My interest is how to accurately and simply communicate information about crime, disorder and police activity to tackle it, to the public. The purpose of doing this is to make people safer and to increase confidence in the police. The latter should encourage co-operation with the police and compliance with the law.

The more I have studied the subject of the public's perception of crime and the police the more I am convinced that the key to improving both is for the police to communicate to the public at a local level what the problems are and most importantly what they are doing about.

Just before Christmas I was asked by a senior detective whether he should inform local residents by leaflets about drugs raids police had carried out in their neighbourhood. He was concerned that not all the raids would be a success and he would reduce the confidence in the police when this happened. My reply was along the lines of "Do the people in your borough think there is a drugs problem? does it reduce their safety and increase crime? are they worried about how it may affect their children? do they think police should be doing something about it?" Assuming the answer is "Yes" then the public should be pleased by the police activity as long as the leaflet carefully explains police procedures to ensure the right places are targeted and the reasons why success it not always achievable. Along with an expression of determination to tackle the problem and an anonymous phone number they could ring if they had useful information.

With the overarching Performance Indicator (PI) for police in England and Wales being "Increasing confidence in the police" it is important that the police are not only working on those things that matter to the public but they are perceived to be doing so. The art of communication is the means.

More Poisson

Following on from my last post I have carried an analysis of the grid square that had the most street robbery allegations recorded on the MPS crime recording system - CRIS. In the financial year 2007-8 45 allegations were recorded giving an average of 0.9 allegations per week. I allocated each allegation to the week in which they occurred.

20 weeks had no allegations
21 weeks had one allegation
11 weeks had 2 or more allegations

This results in the percentages of;

0 = 38.5%
1 = 40.4%
2 or more = 21.2%

The expected values for a Poisson distribution are;

0 = 40.7%
1 = 36.6%
2 or more = 22.8%

Therefore the answer to the question did the occurrence of street robberies from week to week in a specified 250 metres by 250 metres grid square in a London Borough in financial year 2007-8 follow a Poisson Distribution is "Yes".

Poisson to digest at Christmas

I have had a few things to think about whilst digesting the Christmas fare.

Dan asked whether the data I presented about street robbery's likely victims and likely offenders is a Poisson distribution. My answer is "No it is not" because the whole point is that certain people are more likely than others by their demographic (and probably their geodemographic (we have yet to come to that)) characteristics to be victims and/or offenders. This means that the distribution is not random. The problem is it is not deterministic either therefore it can be argued that there is a degree of randomness.

To discuss this more practically time and location need to included as these are the other elements in the commission of a crime (a commodity to steal is possibly another element but in the present discussion this is assumed to be part of the victim's characteristics). I have shown that in the brief analysis of time and place that these factors appear not to be random either. There are specific times and specific places where crimes are more likely to occur than other times and places. So all four elements for a crime to occur are not random.

If on the other hand if a question is formed in such a way that the non-random elements become insignificant it may be possible to use a Poisson calculation or calculator to determine the likelihood of a crime occurring. For instance, if it were known that in a particular grid square an average (mean) of one street robbery occurred every week what is the probability of none, one, or two or more occurring in any one of the weeks? Using the calculator the answers are 37%, 37% and 26% respectively.

This is interesting. The degree to which the location, time, offender and victim are not random can be tested against a theoretical Poisson distribution....................

Location, Location

Continuing with the street robbery analysis in a London Borough. I am experimenting with a minimalist visualisation of the data to show how two different but connected police information systems record the location of the occurrence of the same crime type.

The top diagram shows the locations of robberies recorded on the MPS crime recording information system CRIS. Each robbery location is plotted as a point. This is then overlaid with a grid with squares measuring 250 metres by 250 metres. A spatial join of the two layers is carried out which counts the number of points (or robbery allegations) in each grid square. The grid square is coloured according to the legend with dark colours representing squares with higher number of robberies.

A robbery allegation can be made to police in a variety of ways. One of which is by telephoning 999 if the robbery has recently occurred. Police treat such incidents as emergencies and respond immediately. These incidents are recorded on the MPS Computer Aided Dispatch (CAD) system. Not all calls result in an allegation of robbery being recorded on the CRIS system. The middle diagram shows police emergency robbery calls plotted in the same way as the CRIS allegations.

The bottom diagram is showing the two diagrams above merged with the layers on top of one another to show the similarity and differences in the locations.

Likely victim and likely offender

I will discuss theories about the occurrence of crime in more detail in later blogs because they explain non-randomness, variation and a degree of predictability of where, when, to whom and by whom crime is committed.

To me the most important element in whether a crime is committed at a time, at a location and on a victim or commodity is the existence of a "likely offender". Now "likely offenders" differ with different types of crime. This is logical when you think about it, you would not expect a con-artist to be the same type of person as a street robber.

Profiles of likely offenders can be quite specific. Just using age and gender allows a relatively small percentage of the population to be identified as likely offenders for street robberies in the borough under analysis. (see graphs). Interestingly there is a matching profile of likely victims

Differences in age and gender

Following on from yesterday's brief analysis these graphs show that the street robberies after school appear to be a problem for boys rather than girls. It is possible to gradually build a risk profile for people based on simple demographic characteristics.

Starting to bring it together

Starting from today I am going to post data relating to street robberies in a London borough. It is easy to keep the borough unidentified in graph form but less easy in map form. I will see how I do. The data is deliberately historic. Its source is the initial classification of robbery allegations recorded on the Metropolitan Police crime recording information system CRIS. In recent posts I have touched on time and yesterday I introduced demographics and geodemographics. This graph shows the demographic of age against time. It clearly shows that robberies which have victims who are under 20 years of age occur at different times to those people of older ages. The times suggests school going home time as the vulnerable time for the lower age group. This analysis is reinforced by a graph of Saturday and Sunday robberies.

Now I have to map these robberies out to see what geographical relationship, if any, they have with schools.........

An introduction to Geodemographics and a little bit more

The word geodemographics is a real word that is not recognised by spell checkers so I think it requires some explanation. Demographics on the other hand sails through with no problem. Demographic data are of course details or characteristics of human population. The "geo" in front adds location to the demographic data to create the simple definition that geodemographics is the analysis of people by where they live. Commercial companies use it to sell their products. For instance, if you are selling lawn mowers that you can sit on and drive whilst mowing the lawn it is useful to know where the people who are likely to have large lawns live. Equally if you sell coffee you want to place your outlets where people want and can afford your coffee. A tip if you are thinking of opening up a a coffee shop (or any other type business) open it it next to a like business that is doing well, that's where your customers will be, not where there is no competition.

With the magic of a google map search you can find the locations of various chains of outlets. I have tried Starbucks, Nero Coffee, Lidl, Waitrose, Pound Shops, Loch Fynn, Carluccios, Pizza Express and even Polish Shops to discover the offices in London and the richer and poorer areas of London based on the cliental these companies are trying to appeal to. Now I need to find proxy identicators of variations crime and disorder.

The Art of Mapping

My shared office at UCL is right next to the Slade School of Art though I have yet to visit. One of the pleasures of entering a new function room at UCL is the opportunity to be able to study the paintings hanging on the wall by former students from the Slade. Augustus and Gwen John, Stanley Spencer, William Orpen, Paul Nash and Ben Nicholson are some of the more recognisable names.

I learnt this week that my colleague and fellow blogger James has a love of snow so its been a good week for him. I on the other hand have had a love of art from an early age, not something that was understood or encouraged in a family of scientists. And contributed, with other things I think, to me joining the police.

Art is about all sorts of things but the most important is the connection with conscious and sub-conscious emotion and the conveyance of a message or an idea. Good art does this. The changes and revolutions in art over the ages have been about the techniques by which this is achieved, not the raison d'être.

Techniques in art are about how reality is represented or deliberately misrepresented. Artist try to tackle the problems of dimensions, time and movement within their art.

Geographical Information Science (GISc) includes geovisialisation which is an area in which science interfaces with art. It graples with same problems - dimesions, time and movement and the representations of reality; it tries to tap into the intuitive mind. I have been trying to explore new ways of mapping crime and police related data. The problem I am finding is the representation of time together with location in a simple and understandable way.

Back to Art and the Slade. Eugenie Scrase, a current student, has recently won the Saatchi Competition on BBC2 for her piece "Trunkated Trunk" which is being exhibited at the Hermitage, in St Petersburg, Russia. (see the picture above) This is good art. There is movement, there is impact, there is history, there is the a reversal of 3D to 2D, there is Fibonacci's spiral and there is purity.

Trust in Crime Statistics

The Reassurance Gap (see last post) had repercussions not only for the police but crime statisticians which are still being addressed. The UK Statistics Authority published an interim report "Overcoming barriers to trust in crime statistics" yesterday. A copy can be found here. It is all pretty dry stuff but of great interest to us who are studying public perceptions of crime and the police - in a similar way I suppose a publication about a misunderstood pond weed would be to certain bio-geographers.

One of my favourite quotes is "The police are more trusted as a source of information than civil servants or politicians."

And without being too flippant that is the gist of the problem. Crime stats in England and Wales have become too politicised and that is the reason they are not trusted. That is the clear message within the report but not one that is properly addressed by the report. Could it be because it is written by civil servants for politicians? There is a high potential that meddling could make things worse. Paragraph 104 does recognise this;

"People’s perceptions about crime are influenced by the anti social behaviour they observe locally. If the notifiable list were to be reduced (paragraphs 75 to 83), then even less anti social behaviour would be reflected in the national crime statistics. Distrust of the national figures might therefore increase."

The obvious conclusion is to include more anti-social behaviour stats and exclude less relevant ones but that does not seem to be even considered.

I will no doubt write more about this report in later blogs I am in danger of going into a rant if I write more now.

Happier Police

Reassurance Policing was introduced as a policing style to police forces in England and Wales to address the problem of the Reassurance Gap. This Gap was the difference between what the official crime statistics showed and what the public perceived to be the case. The public thought crime was rising despite the fact that recorded crime was falling. The Association of Chief Police Officers in England and Wales (ACPO) published a report in December 2001 that outlined the seriousness of the Reassurance Gap to police and how it should be addressed. The solution was to introduce dedicated neighbourhood police teams who would patrol on foot. The report can be summed up in one sentence;

"In short police forces need to demonstrate visibility, accessibility and familiarity and these are the main elements of this report."

In an early draft of an academic paper I wrote "police needed to become more visible, accessible and familiar" only for my supervisor to cross it through and comment " I do not want familiar police." I think he had visions of police officers going round calling everyone "mate" and "luv".

Anyway Sir Paul Stephenson the Commisioner of the Metropolitan Police has recently commented that police officers in his force are now happier. It nice to know in this Christmas season.

At least now when a police officer calls you luv or mate they should be smiling or perhaps even laughing.

BBC setting the standard

These are images from a new BBC interactive website launched today. It can be found here.

I am impressed with it because it allows the user to find information that is relevant to them in a simple and quite specific way. And because of that it is much better than the crime maps that are currently produced. Note too that there is point rather than polygon information.

This is an excellent template for crime maps.

Does time have at least two dimensions?

This entry to my blog should be seen as thinking out loud (tol), work in progress or leaveit out lad (lol).

Police incidents, crime, police patrols and other police activities can be mapped because they happen at locations in space and geography. These locations are described by two co-ordinates - x and y, longitude and latitude, easting and northing, etc. Location therefore has two dimensions. If height is added there is a third dimension. Spatial statistics are based on analysis of distances between objects in real or conceptual space. Spatial auto-correlation is based on the observation that objects that are close to each other share similar conditions. This is true of weather, house prices, health to name three. Crime and police incidents are strongly spatially autocorrelated. I will explain why in later blogs.

Now time is traditionally seen to be linear and one dimensioned. There is also temporal auto-correlation based on the fact that events close in time share similar conditions. One of my colleagues in UCL Centre for Advanced Spatial Analysis (CASA), Fabian Nuehaus recently talked at a CASA seminar and reminded me that life is full of rhythms and cycles. That means that for instance Sunday morning is closer in conceptual time space for autocorrelation purposes to next Sunday morning than the next morning (Monday, as long as it is not a bank holiday). Equally 6pm in closer in rhythm space to the next 6pm than the next 6am. This means that time has at least two dimensions for analysis purposes. Can we map it and use spatial statistics to analyse it?

Policing Pledge

These are the two little clips advertising the Policing Pledge. Within the PP there are commitments about crime maps but the Metropolitan Police choose to emphasise their commitment to responding to calls, neighbourhood patrols and supporting victims of crime in their literature.

Safeguarding of Children and Information Systems

I approach this subject with some trepidation because whatever I write may be subject of misinterpretation as crimes against children is understandably a highly emotional issue. I feel I should though as fear is a motivating factor in the introduction of preventative measures and information systems are increasing seen as the solution. And because it is again headline news in the UK.

Again two different approaches are highlighted. In the United States "Megan's Law" allows a map based search to be made on neighbourhoods to find out the addresses of people convicted of sex offences that involve children or are otherwise serious. In the UK the approach is to only allow adults to have contact with children outside a family and friends environment if they have had their details checked by police and no concerns have been raised. This website goes into the details.

I just want to make one quick observation today.

There are serious social consequences that result from either approach. Megan's Law must push those affected to the margins of society. Police checks have the consequence of treating all adults, especially male ones (but not exclusively after the Plymouth case) as guilty for wanting to have contact with children until proven innocent. This discourages male adults from becoming involved with children's activities. I used to lead a youth group for a number years on a voluntary, unpaid basis. I would wonder if it was worth the hassle and suspicion now. The point I want to make and will continue to make throughout this blog is that crime prevention often produces unwanted side effects. It is the responsibility of the policy makers and the police to maximise the positives and minimise negatives. I am not convinced that the balance is properly being found in the UK and I am uneasy about the US solution. A major part of my argument is - do either work in the way they are intended? or are either or both a political rather than a policing solution? But that is for later blogs.

Wonderful Maps

My background is policing and information systems. So the mapping side of things is new to me so I am making new discoveries every day. My colleague, Dan directed me to an impressive Google Map mash-up of shipping round the world in his blog. As he has been kind enough to show an interest in my meanderings I thought I would investigate a crime angle on shipping and came up with the obvious of the Somali pirates. This led me to a United Nations website of wonderful traditional pre-google mash-up type maps. There is a whole series of maps on the Somali pirate problem and how it has evolved, expanded and been displaced over the years. All the maps are good but I think this one is a masterpiece. There is even a spatial analysis document. If only the UN could be as good at enforcement as they are at mapping.................

I decided not to investigate Dan's other link further.

Point or Polygon

Following from my post yesterday in which I introduced the subjects of privacy and purpose. I will provide two links to illustrate the different approaches to these considerations. The first is from London and the second is from Los Angeles. Both maps are police generated for the public.

http://maps.met.police.uk/
http://www.lapdcrimemaps.org/

The London map is based on polygons to protect privacy, the LA map is based on points with information about time and place.

The LA map provides far more information which the viewer can use to be safer. In my view the policing purpose of public safety and crime prevention has been wrongly been superceded by privacy in London.

I will of course return to this subject because there is much more to be said.

Purpose and Privacy

Just a quick post today to introduce the important issue of privacy and the mapping of police data. This is where the importance of purpose is relevant. If the maps are for the purpose of police accountability, performance indicators and the like (ie political) then there is little justification to compromise individuals' privacy. If on the other hand the maps have a policing purpose like keeping people safe and crime prevention and investigation then there are plenty of precedents where privacy is trumped in the hierachy of priority.

Crime, fear and mapping

View Richmond Homicide Map 2009 in a larger map



I was originally attracted to this map through this link;

http://richmondconfidential.org/2009/11/30/crime-overlays-poverty/

This simplistic analysis actually raises far more questions than answers but it got me hooked for half an hour wondering what sort of place Richmond California is. As an outsider it makes me feel that Richmond is a dangerous place. But is it actually? I have no point of reference. To a large extent I am a web voyeur, the serious incidents do not affect me in my everyday life. What is important is how does it affect the people in Richmond. Does it increase their fear of crime, does it make them avoid certain places, does highlight a problem that needs fixing, does it reassure people because it shows that people like them are not usually victims?

This type of information being in the public domain, though challenging, provides impetus for the authorities to address the problem and individuals to decide how they and the ones they care for do not become victims. The outcomes should be positive but if not managed by the police correctly can lead to further segregation, more guns and a fortress mentality.

The first thing to say about this map is that appears to be created by a journalist rather than the police. I think it conveys a powerful message of a serious problem with guns in Richmond California, which if you pan the map out is just north of Berkeley and San Francisco.

Prime Minister's Speech on Smarter Government 7/12/09

As I hinted yesterday crime maps are political because they seem to provide the right sound-bites. The White paper on policing (see the link on my 6th December post) mentions crime maps over ten times. The Prime Minister mentioned crime maps in his speech on smarter government.

"Through the new online crime maps which went live last month (see link in yesterday's post)- allowing for the first time everyone in the country to search by postcode for facts about crime in their area and what is being done about it - we are exploring how people can use police data on late-night incidents to help them choose the safest routes home and to post travel tips and security tip-offs for others."

This is interesting because the crime maps he refers to do not even come close to doing the latter and only partially achieves the former. I will try to explain this over future posts as I slowly plod along.

For cyclist out there and I know there are plenty who may read this he also mentioned cycle accident maps as a shining success. I think he must be referring to:

http://labs.timesonline.co.uk/blog/2009/03/11/uk-cycling-accidents/

Again I think the sound bite may be more effective than the reality.

If you want to read the full speech this is the link. http://www.number10.gov.uk/Page21633

What is the purpose of Crime Maps for the Public?

I will leave this question hanging in the air because I do not have a convincing answer for the crime maps that are currently being published on the Internet.

In the UK crime maps for the public is a development that has been driven by politicians rather than by the police. It would therefore not be unreasonable to suppose that the purpose of the maps is more political than policing.

http://maps.police.uk

See what you think.

Government White Paper

Even though the headlines proclaimed the slashing of overtime budgets and general cost savings of £500 million by 2014 the Government White Paper "Protecting the public: supporting the police to succeed" published on 2nd December 2009 is actually good news for the police. It re-enforces and further consolidates the Reassurance/Neighbourhood Policing style that has gained supremacy in the last few years. A copy of the document can be obtained at;

http://police.homeoffice.gov.uk/publications/police-reform/protecting-the-public?view=Binary

It is good news for the police because even though police forces have to save money the policing style actually is more resource intensive than the alternative styles of policing. In many places in the USA the type of Neighbourhood Policing that is now practiced in England and Wales is considered unsustainable due to cost. At a time when government needs to save money a trimmed down police service is an option. An option that the Conservatives are no doubt looking at very carefully.

Introduction

Perhaps it will be helpful if I briefly explain who I am and why I am writing this blog.

I am a retired police officer from the Metropolitan Police Service (MPS) in London having done my 30 years. I joined on 20th November 1978, my mother's birthday and retired on 5th December 2008, father's birthday. So as of today I have been retired exactly a year. I was fortunate enough to successfully apply for a MPhil/PhD 3 year research project at the Department of Geography at University College London (UCL) funded by the Economic Social Research Council (ESRC) and my former employee, the MPS as part of a CASE award ( I have yet to discover what CASE stands for). The title of the award is "Real time geodemographics for reassurance policing and crime prevention."

I have spent my first year trying to work out what the title means and to reduce it down to a PhD thesis. Now that I have done alot of reading and learnt a few new skills, I have enough confidence to presume to think I have something different and hopefully original to write. I will obviously be also relying heavily on my police knowledge and experience.

First

This is my first ever blog so I will start as I intend to go on by not worrying too much about everything being perfect but just getting my thoughts, observations and possibily rants down on paper. But of course it is not paper, it is a multimedia, multiconnected machine world that I am entering. It is how police can make the best use of that technology to convey their messages through maps and other means to make people understand what dangers are out there and how they can be safe and equally importantly feel safe that is the theme of this blog.