Qualitative data analysis using MS Copilot Chat

Overview
The system was used to support analysis of consultation responses collected from government departments as part of an assessment of compliance with Human Rights Minimum Standards and Minimum Core Obligations (MCOs).
How it is used
The exercise involved reviewing and synthesising a large volume of qualitative information across 9 rights areas, 63 MCOs, and three assessment questions per MCO. The primary objectives were to increase efficiency, reduce the time required for manual synthesis of qualitative data, improve consistency in the assessment of responses, and enable more sophisticated analysis than would have been feasible through manual review alone.
Scope and Limitations
The system was designed to create a number of reports or chapters that summarise the responses under different criteria and questions and to identify equalities references.
The system was not designed to infer further learning from the qualitative data provided. It was not intended to make policy decisions or recommendations.
Human Oversight
Throughout the process, outputs were proved and quality assured, decisions on format, language drafting etc remained with policy colleagues, and extensive editing was necessary.
Some data inputs were not handled well, especially when very large data sets were processed so manual data handling had to be done for one of the 9 reports. Overall however, the approach did save time and made it possible for the work to be undertaken given the volume of other work and competing priorities in the team.