Methodology
The methodology underpinning Neighbourhood Knowledge Management (Δnkm) is founded on household level data typically found in local administrative systems and public registers. Using the approach, issues can be identified and quantified to a far greater level of detail and variety than is possible using other methods to a geographical scale of the user’s choosing.
In essence the techniques involve matching data from different organisations and sources and then using it to analyse a range of issues at the interfaces for example between education, health and crime. There has been a range of applications to date, such as bids for funding and several new applications are being developed all the time. The techniques have been used for example to analyse the interrelationships between domestic violence, low-income families, social housing, drug offending, noise complaints, and mental health. Outcomes included a detailed quantification of risk factors by neighbourhood and detailed cellular maps of individual neighbourhoods (500 metres x 500 metres) showing the concentration of key risk factors.
The method has many advantages over alternative data sources. These may not provide information within the ideal time sequence or timeframe. Household data offers much greater flexibility in terms of analysis and spatial scale, and because the data are administrative they are usually more up to date and accurate than official statistics. There are geographical mismatches due to boundary and administrative changes. The alternative sources that have been considered are:
1. Census and other survey statistics: these are expensive and time-consuming to collect and analyse. They often have limited information particularly around patterns of service use. The Census is collected every ten years. This means it cannot document trends that occur within, for instance, the lifetime of the Neighbourhood Renewal programmes. There is concern that Census figures do not reflect reality. There is a general lack of regularly produced information at the geography of local area partnerships (LAPs), Surestart areas etc. Frequent boundary changes make the accurate monitoring of change almost impossible. These constraints also contribute to the poor diagnosis of problems that impact on the effectiveness of policy and resource allocations.
2.
Area-based analysis of service provider data: postcode, enumeration
district or ward analysis is potentially very misleading because it assumes that
statistical associations are ‘real associations’ – the ecological
fallacy. This can lead to flawed conclusions.
Types
of application
Technical services
Between them local service providers such as the PCT, local authority and
police hold and maintain at least 20-30 data sets that are currently underused
used for local analysis and planning. Where they are used they are not shared
across agencies or even between departments within the same agency. Much effort
is wasted on trying to adapt other data official data sources, which are frequently
out of date, definitionally incorrect, or apply to the wrong geographical boundaries.
With successive applications of Δnkm in different area we have developed
a range of technical services to help unlock the information contained in these
data sets. These include the following:
Designing and delivering Δnkm projects
Our experience is that users prefer to start
by applying the techniques to one or more key policy areas that potentially straddle
several service providers. Usually a lead department in the PCT, Police or local
authority will take the lead and help steer the project. Most projects do not
take long – typically a few months. The lessons learned as well as the data
created can be applied in future applications relatively easily. To assist technology
transfer (conserving the learning locally) is useful if on or two key analysts
employed by these organisations are involved in any project or projects.
It is possible to envisage three levels of application:
Basic package
In the basic package we can take 4-5 data sets, clean and match them to the
local property Gazetteer. These could include a GP register, electoral roll,
council tax and the school pupil roll. The data sets are anonymised and geo-coded
and then analysed according to the application required. This dataset allows
us to produce an up-to-date and accurate population figure, analyse socio-economic
indicators such as poverty and ethnicity, access to primary care and education
attainment and provide detailed neighbourhood maps to any specification. Analysis
is always in aggregate form but can be done for any sub-district unit or population
group required as long as individuals are not identifiable as a result. No
personal information is ever disclosed.
Advanced user
In the advanced user package we employ more data sets and extend the range of possible applications. The additional data sets could include crime, housing stock condition, hospital admissions, and mortality data, data from social services and so on. The level of analysis required and the techniques used are more involved and therefore the cost is higher.
Expert users
For expert users we add further data sets and employ additional analytical techniques. The analysis is often directed at a specific problem or policy area. Considerable effort goes into identifying the problem, working closely with policy customers, adapting the approach to the issues and problem solving. The outputs usually include policy advice, reports, training, databases and so on.
Please contact info@nkm.org.uk for prices.
If user
needs are continuous or the work is planned to occur in stages, then subsequent
applications can be carried out for a smaller sum, as the work carried out in
earlier work often feeds directly into new work.
