NIM and intelligence
NIM and statistics
Dysfunction of NIM
The role of Principal Analysts
Legal Status of NIM
What then must be done?
Copyright notice: text from this website may be reproduced freely, so long as IntelligenceAnalysis.net is quoted as the source
About this website
The purpose of training is to teach skills; for instance, a course in bricklaying teaches how to lay bricks. In the context of intelligence analysis, training teaches how to develop actionable intelligence. Intelligence analysis, like brick-laying, is a practical skill that can be acquired by many people.
No training course in the United Kingdom teaches intelligence analysis. Instead, courses promote National Intelligence Model. For instance, the National Police Improvement Agency states that it is " dedicated to supporting the implementation of the National Intelligence Model". Kent Police ensures that course participants will " understand the key features of NIM and its need for analytical products". And Yorkshire Police provides students " with the skills and knowledge necessary to function as an intelligence analyst in a modern NIM environment".
By teaching National Intelligence Model rather than practical skills, training courses teach an agenda based on a particular set of attitudes. As such, these courses may be expected to display certain characteristics: that no distinction is acknowledged between analysts as investigators and as statisticians; that analysts are not recognised as investigators in their own right; and that managers control - rather than manage - the intelligence analysis process. These characteristics are evident in all UK intelligence analysis courses.
This page examines five training providers: RISC; NIAT (Manchester, Yorkshire); NPIA; the Jill Dando Institute; and Skills for Justice.
Research and Intelligence Support Centre (RISC), Wyboston, Bedfordshire
RISC is a commercial provider of "tactical and strategic level" training, and says that the nine analytical products described in National Intelligence Model comprise the "toolkit used by Intelligence Analysts". Thus, from the outset, RISC makes no distinction between which products are investigative and which are statistical; all nine are deemed the work of 'intelligence analysts'.
RISC offers two main courses. Its Strategic Assessment and Analysis Course introduces students to " new thinking skills, for example, interrelationship digraphs; futures wheels; and triangulation factor analysis futures cones", plus "statistical interpretation as an aid to the development of strategic intelligence products, including: statistical perspectives; advantages & limitations to statistical applications". None of this has anything to do with intelligence or investigation, but it does show how the term 'strategic intelligence' is used as a euphemism for statistical analysis.
RISC's second course Intelligence Analysis is also mainly statistical in nature. The principal point of reference for analysts taught by the course is not the criminal but various categories, such as localities of crime. Accordingly, the course teaches interpreting statistics, standard deviations and mapping. Having estabished the intelligence analyst as a statistician, the course then contradicts itself by teaching analysts how to do link charts; but these are not taught as analysts' investigative tools, but as products for "audiences" and "clients" who have tasked the analyst to produce them.
National Intelligence Analysis Training (NIAT, a consortium of several forces, including Manchester and Yorkshire)
Manchester runs three courses under the heading of intelligence analysis training, the first of which is NIAT Initial. This is said by the course summary to "...explore the concept of Criminal Intelligence Analysis"; yet, it is also described by the summary as suitable for Analysts performing, "either intelligence, strategic or volume crime functions within law enforcement". Since at least one of these functions is code for statistics, the claim that the course will "explore the concept of Criminal Intelligence Analysis" is false.
The next course Crime Pattern Analysis is entirely statistical, as indicated by the course summary:
The third course NIAT Strategic Analysis teaches, " specific techniques and methodology to enhance the content and relevance of strategic assessments to support control strategies at all levels". Again, not a course concerned with investigation and intelligence.
Thus, of three courses claimed by Greater Manchester Police to relate to intelligence analysis, two don't; only the Initial course refers to it, and this includes a statistical component. GMP further compromises its training by saying that people who take NIAT Initial can "progress" to, and "specialise" in, the other two. In other words, someone who completes a course, nominally in intelligence, is expected to evolve into a statistician.
The Yorkshire Strategic Assessment and Analysis Course, like that of Manchester and RISC, is for staff "conducting Strategic Analysis and engaged in the production of a Strategic Intelligence Assessment". It is management information, not intelligence or investigative.
Yorkshire's Crime Analysis Applications is a confused mixture of statistics and intelligence. The course summary follows, with intelligence highlighted green, and statistics red:
National Policing Improvement Agency
NPIA training for analysts on major incidents has been referred to elsewhere on this website; this is where the stated aim of the course is to train the intelligence analyst to work within 'agreed frames of reference'. In making this declaration, NPIA is shutting the analyst out of the investigation. NPIA expands on this with the following course objective: "Explain the role of the analyst as a professional advisor to the Senior Investigating Officer during a major incident":
It is not immediately clear why NPIA should regard intelligence analysts as 'advisors' when deployed on major incidents (we can dispense with the prefix 'professional' since the analyst is hardly expected to be unprofessional). The role of an analyst on a major incident is not significantly different to that on a non-major incident, since the intelligence analyst is performing the same essential role of developing actionable lines of enquiry. Nor is the role of senior investigating officer (SIO) significantly different to that on a normal crime; a major incident will need more detectives, so one is appointed to take charge of the rest but, otherwise, there is no fundamental difference. The incident room has its own database, more staff, is mainly statement-based, but it is still an investigation with the same processes and objectives as any other crime investigation. Yet, NPIA has taken it upon itself to define analysts on major incidents as 'advisors'.
Describing the intelligence analyst as an advisor infers that the role is to provide 'advice'. This is incorrect; the purpose of the intelligence analyst is to provide intelligence, preferably actionable. Whether the intelligence is actioned is a matter for the SIO, but it does not follow that the analyst is an 'advisor' or that intelligence is 'advice'; intelligence represents the 'available truth' and should be acknowledged as such.
To illustrate this in another context, the SIO's role is to produce evidence for counsel to turn into a prosecution. The SIO is not regarded as an 'advisor' to legal counsel, because legal counsel does not need advice. They look to the SIO for evidence, not advice. So too, the SIO looks to the analyst for intelligence. NPIA's redefining of the intelligence analyst as advisor implies that the analyst is not part of the investigation but outside, offering easily ignorable advice to those within.Jill Dando Institute of Crime Science
The Jill Dando Institute has six analysis courses, all of which are statistical, yet are described as suitable for "intelligence analysts". The statistical nature is demonstrated by the following quotes from the course summaries (link):
How to Become an Effective Problem Solving Crime Analyst: "... directly compatible with the National Intelligence Model ... geographical crime displacement; measuring crime switch displacement Distinctions between different types of hotspots; diagnosing hotspots, and understanding how to use volumes and rates analysis techniques for understanding long term change; seasonality; crime patterns by day of the week and time of day ... statistics used in outcome analysis".
Getting the Right Stats and Maps for Today's Policing: " statistics for characterising data sets (e.g. mean, mode, variance and standard deviation) and an overview of the principles of inferential statistics produce solid descriptive statistics, using crime-based examples Analysis of spatial and temporal trends in crime data, including methods for investigating daily, weekly, monthly and yearly trends and considering the role of seasonality exploring the role of inferential statistics including chi-square, correlation statistics and logistic regression".
The next two courses can be dealt with together. Understanding Hotspots and Advanced Hotspot Analysis: "GIS-based mapping, hotspots, statistics, crime patterns, Geographic boundary thematic mapping, Quadrat thematic mapping, Continuous surface smoothing Kernel density estimation hotspots, spatial statistics, spatial significance, spatial autocorrelation, spatial association, Local Indicators of Spatial Association, dual kernel density estimation determine areas that can be statistically defined as hot from those that are not, plus rank each hotspot according to significance threshold (i.e. 95%, 99%, 99.9%)".
The final two courses: spatial regression, involving coefficients, Chi-square and linear regression, spatial lag models and geographically weighted regression, and neighbourhood analysis covering geo-demographics.
None of the above has anything remotely to do with intelligence, despite the claims made by the Jill Dando Institute to the contrary.
Skills for Justice
Skills for Justice produces ten courses on intelligence analysis. Each of the course summaries refers to intelligence analysis and covers a different aspect. Rarely, however, do the summaries state what the courses are about. For instance, the course summary for Unit 9 says "how to analyse information" but it does not state whether the word 'analyse' is being used in the investigative sense or the statistical sense.
The truth slips out in the course summary for Unit 3 on applying analytical techniques: "This unit may apply to analysing the activities of individuals or analysing patterns and trends". Thus, both functions are taught as one course, the implication being that one person is expected to do both.A consequence of this malpractice is seen as Unit 1 which requires the analyst to meet the following "performance criteria":
Before attempting to explain what is wrong with the above, here is what an intelligence analyst actually does:
Note the difference between the two approaches. According to the Skills for Justice, the intelligence analyst undertakes work only after extensive and detailed discussion on 'requirements', 'techniques', 'objectives' and 'parameters'. Thus, the intelligence analyst is both subordinate and controlled by a police officer with whom he must 'negotiate' and obtain 'agreement'. In the second approach, the analyst conducts their work and then discusses the analysis with the detective.
In both instances, the intelligence analyst works with an investigator, and in both cases the investigator is entitled to influence and raise issues with the analyst, but only in the second approach is the intelligence analyst is assumed to be competent in their own right. Consider the matter in a different context: when a detective is allocated a crime, he is not expected to say to the crime desk: "But what do I do? What are your requirements; shall I explain my products?" The detective is expected to know what to do, not wring his hands. Similarly, when a SOCO is assigned to a case, they turn up, examine the scene, and write their report. They do not sit down and explain their techniques, negotiate objectives, seek permission and otherwise apologise for existing.
How can Skills for Justice make such a basic mistake? To understand Skills for Justice's inverted perspective, we must return to the logical difference between statisticians and intelligence analysts, explained here: statisticians control their questions; intelligence analysts do not. Everything advised by the Skills for Justice assumes that questions can be set at the beginning; this is the practice of a statistician, because the statistician's point of reference is the client who represents the perspectives on which the statistics will be based. The intelligence analyst, by contrast, is geared to the answer (the criminal) and so does not rely on the detective for frames of reference, for the detective has none to offer other than the crime report, which the analyst can read for themselves.
What we learn from this last example is that training which is appropriate for a statistical analyst is not appropriate for an intelligence analyst. This single, simple principle is not understood by any of the training agencies listed here.
See also: the OTHER problem with intelligence analysis training and recruitment - Telex entry 4 January 2011
|Next > dysfunction of NIM|