Throughout a diagnostic, local areas rely heavily on their data to identify parts of the local system to focus on and give them an understanding of how they can aim to tackle challenges in these areas. The more data you use, the richer the evidence base you have, enabling you to make more informed decisions and the most effective change. As a result, the gathering and setting up of data and sharing agreements is crucial to delivering a successful diagnostic.
Context: The way data is collected and stored can vary significantly between local areas. In this case, the DBV authority found difficulty in drawing from multiple data sets across different case management systems.
What they did: To overcome this, they categorised the data they needed for the diagnostic and put together a small team to collect it into one place. The list of datasets on the right demonstrates their approach and the data categorisations they used. The local authority also found it useful to consider what data had already been used for SEN2 returns as a starting point.
What they found: This method quickly identified where there were gaps in the data that might prevent them from being able to carry out the initial analysis. Once these gaps were filled, they could spend time on gathering more detailed information. They later found that individual data points such as specific placements, primary needs and EHCP start dates were crucial to help inform the priority areas in their diagnostic investigations.
There is also no doubt that this is a lot of data, and aggregation can often be a complicated and laborious task. However, by completing it thoroughly at this stage, the local authority reflected that this meant there was less demand on their time and resources during later stages of the diagnostic.
Data validation is a vital step to avoid rework and prevent drawing misinformed decisions later down the line. The data you collect at this stage will be used throughout the diagnostic, so it is vital that it is collected accurately and check for inconsistency.
Context: Similar to the first case study, the quality of data can vary both year by year and between different local areas. For instance, during the set-up phase of the DBV programme, a local authority noticed that they had considerably more data points in 2021, whilst there was lots of duplication and several errors in others. This meant they had to find ways to reach the same level of data in other years.
What they did: In this case, the LA’s data team worked through the tables of data they had created, ranking each one based on any inaccuracies or missing information – see Table 1 as an example. By keeping a log of changes and updates, they were able validate the data throughout the gathering process and catch mistakes early, rather than testing the entire dataset at the end.
Key considerations: Some common questions to bear in mind when testing the accuracy of your data might be:
