Click here to go directly to the CWIP App Version 2 Beta. For a brief overview, read on…
It’s been a year since I started working on the first version of the Child Welfare Inequalities Project App. The purpose of the app was to help social workers, policymakers, local government, activists - anyone, really - with the task of representing trends and patterns in, and pulling out basic statistics from, children’s services data. Anyone familiar with the statistical releases of the Department for Education will know that they are not always the most user-friendly files, they can be hard to combine, especially over time, and it can be hard for people with limited technical or statistical knowledge to extract the information they want from them. It is even harder to turn the data that’s extracted into something engaging, like a data visualisation, and this may take much more time than an activist or advocate might have. Quite frequently, the data is archived or requires some transformation, making comparisons over time difficult.
The CWIP App brings together around 16 different datasets at the local authority and neighbourhood level, with a range of features including being able to adjust for deprivation effects using the results from linear models calculated ‘on the fly’; the ability to compare multiple different graphs side-by-side; and the possibility to create literally hundreds of thousands of possible graphs. Moreover, in my opinion, these graphs look beautiful (while being informative as possible), and all have a consistent theme for non-jarring implementation into any organisations’ reports. All of this is done through a UI and hosted online, which means that nothing needs to be installed and no programming knowledge is required. The CWIP App is completely free, and has been funded by the University of Sheffield. All you need is an internet connection and a computer. You don’t even need to register for an account.
Version 1 of the CWIP App was my first attempt at building an R Shiny application and it was, to be blunt, very clear that it was my first time. It was well received, but there was consistent feedback about its limitations; many of which I didn’t have the technical skills to overcome. The code was full of redundancies (even more than this new one!), slow, and things like data storage and transformation were inefficient. In the space of a year, not only have the R community and Shiny resources grown, but my own experience and practice using it has, including when I developed the Mapping Overlaps Gadget (the features of which are replicated here). The offline and fully featured repository for version 1 included a 4.6GB data shapefile… The entirity of this version is around 0.75GB, while containing a huge number of additional variables, and is hosted publicly on Github, where anyone can download it.
This means that all of the code, all of the data, everything, can be examined, scrutinized, modified, extended, and used for whatever purpose anyone wants. You can use the app and its source code for education, practice, activism or management. And any graph you produce can simply be copy and pasted elsewhere, or for interactive plots there is a ‘download plot’ link.
With that introduction, let’s move onto the practical matter of what features are available…
The Data Dashboard allows you to quickly look at trends in variables from Section 251 returns, Children Looked After statistics, and the Children in Need Census…
You can pick from two different ways of visualising trends: as a simple trend plot, or as a bar showing the proportion of the funding in a local authority against the average in England. You can also get a sense of how these differences look on a map, what the variables’ relationship with deprivation looks like, and which local authorities have the most similar values on average or trends across time, by using the accompanying plot feature.