To address what many have deemed a controversial statistical technique, the IPA has published a new guide offering advice and best practice to tackle the weighty issue of data integration.
Data Integration Explained focuses on how different forms of data integration work, best practice, how to prepare and evaluate the results and what to bear in mind when using integrated datasets.
Through correct integration, the IPA hopes that it will be possible to leverage further insights.
Lynne Robinson, Research Director, IPA, explained that undertaking data integration is a particularly complex task.
“We found this when the IPA began producing the TouchPoints Channel Planner dataset which involves several different types of data integrations,” she said.
“Following this experience, we decided to produce this guide which is designed as a best practice primer in undertaking data integrations of all types so that users can ensure that their merged data sets are as robust as possible.”
Steve Wilcox, managing director, RSMB Ltd, and author of the report, said: “Although data integration is a simple concept, there is no ‘one-size-fits-all’, single best solution and the underpinning statistical theory is necessarily complex.
“I am delighted the IPA has given me the opportunity to help guide users through the key processes and complexities – in my opinion it is as important to understand the limitations as well as the benefits of any approach.”
The report is available to purchase from the IPA’s website.
Lynne Robinson will be writing for Newsline next week as she looks at the project in more detail
