How to turn data into insight

Turning data into meaningful insight is a huge reason to collect data in the first place. Here, we set out the core components and provide some examples.

Acquiring Data

• Look at existing data sources inside your organisation
• Understand ways in which you can use this data (what licences govern and control its use)
• What data can you find elsewhere from partner organisations or even as open data on the internet?

Exploring and Preprocessing

• Reviewing the data and cleaning it
• Ensuring accuracy and usability
• Preprocessing by dealing with unhelpful data or formatting the data so it can all be used together

Analytics

There are many different types of analytics, from proving what has happened in the past (descriptive and diagnostic) to determining what may happen in the future (predictive and prescriptive) the latter can even lead to changing behaviour and therefore outcomes.

To undertake analytics we will combine both structured/unstructured data sources. They have the following traits:

More often than not in healthcare and research we are undertaking statistical analysis. This is effectively some type of analysis that has certified accuracy that the results are correct. The process thereafter would be as follows:

• Describe the nature of the data to be analysed.
• Explore the relation of the data to the underlying population.
• Create a model to summarise understanding of how the data relates to the underlying population.
• Prove (or disprove) the validity of the model.
• Employ predictive analytics to run scenarios that will help guide future actions

Communicating Results

• Visualisation
• Storytelling
• Understanding the many consumer types

Turning Findings into Action

• What impact could your findings have?
• What steps need to be taken to make that happen?
• Monitoring performance against the potential impact to determine the success of the intervention

Analytics are frequently used in medicine and healthcare, and huge success stories exist, such as greater success rates in cancer treatment, as an example. So high-quality data collection, analysis and research remain important for monitoring progress and identifying which initiatives will be most effective at improving outcomes.