Intelligibility measurement is a relatively new science, and to assess intelligibility accurately and at scale is something we are uniquely good at.
Many people have asked what underpins Amplifi, how do we know it works, and what is the impact of simplifying communications?
In this next section, we're going to look at the science and research that underpins Amplifi, and some of the work we've done with Regulators to test and demonstrate our techniques.
The science behind Amplifi
To create the world's first intelligibility assessment tool we've used a mixture of existing and our own unique research, and applied AI and machine learning. We have assessed huge volumes of regulated and legal documents to develop, test and hone our model, so that it is a strong predictor of understanding.
We started with a base of the 50 years+ of readability research, using established models such as Flesch Kincaid and Gunning Fogg as well as other language assessment models and state of the art linguistic frameworks. And we’ve extensively tested the impact of different communications on readers’ level of understanding.
This gives us a way to accurately test how likely comms are to be understood. We can pinpoint and evidence the various factors that are helping or hindering consumer understanding.
Our testing with Regulators and Academic partners
Our research has involved some leading academics in the field of legal innovation and consumer understanding.
We've worked with the University of Nottingham's Professor Richard Hyde and Warwick University's Centre for Financial Innovation on a project in the FCA Sandbox. Here, we tested our intelligibility assessment methods looking at credit disclosures – this included pre-contract information, and default notices. We looked at the impact that simplifying them has for the reader.
When we simplified the debt notice, we showed that you can make them much more understandable. Another result was, as well as improving understanding, simplifying communications also improved the emotional response from readers. They experienced less anxiety and were more likely to take action after reading the simplified notice.
Simplification helps comprehension, but also how the communications are received and perceived by the reader.
In one of our latest projects, our Head of Research Sarah has been working with Richard Hyde again, this time together with the Solicitors Regulation Authority. We'll be co-publishing the exciting results together with SRA in a lot more detail early in Spring 2025.
The impact of simpler communications
In the test we used Amplifi to help SRA simplify some of their regulatory guidance. We tested this before and after to test its intelligibility.
We then put this in front of a wide range of readers. We used Richard's eye-tracking techniques to record how readers engaged with the text. We also refined and applied a comprehension framework to test their understanding.
Importantly, this wasn't just testing what they could remember but also asking them to apply the information they'd read. We effectively tested true intelligibility and therefore the understanding outcome requirements of the Consumer Duty.
The results from our testing closely matched the scoring from the Amplifi assessment.
We showed a significant improvement in understanding for readers who read the simplified copy whatever their level of skill or experience. Simplification benefits understanding outcomes for all readers.
Summary
1) Amplifi uses a unique combination of existing science and new research to test intelligibility
2) We worked directly with FCA and the SRA to test and validate our methodology, showing that it accurately predicted understanding.
3) Simplified information has been shown to help all readers' understanding, whatever their ability, and reduce negative emotional responses