7.2.2: Analysing James’s Multi-Centre Session
Looking across a multi-centre session
- Understand the unique features and challenges of analysing multi-centre session data in RM Compare.
 - Master cross-centre data matching, comparison, and visualisation using LLM-supported workflows.
 - Develop insights relevant to school groups, trusts, or regional clusters, supporting data-driven improvement and collaboration.
 
Introduction & Context
Remember James from the previous lesson (7.2.1)? Well he is now running a Multi-Centre session across 7 schools. You can find the dummy data-set here (James Multi-Centre Session1).
He will want to complete some analysis across the group. At the same time he will want to share data and reporting to individual schools that is annonymised.
Data Familiarisation
Spend a bit of time inspecting the data. You will see again how the session Data Extraction has been combined into a single spreadsheet (your LLM can complete this task remember). Note also how we have added some Meta-Data - in this case from the school Management Information System.
There are a lot of questions we can ask about this data!
LLM-supported Analysis & Visualisation examples
Prompt: "Calculate average/median item scores per school. Use the 'Param as Score' data to do this.
    
Prompt: "Generate a Box and Wisker chart to show score spread and central tendency by centre"
    
Prompt: "Explore cohort differences, outliers, and “peer effect” patterns."
    
    
Prompt: "For the box and whisker create another but only label one school (Riverside) and anonymise the rest. I want to give this to the Riverside headteacher but not name the other schools in this case."
    
Interpretting the results and benchmarking
Prompt: "Which schools stand out for performance or consistency? Are there shared strengths/weaknesses, or notable differences? What interventions or best-practice sharing might be warranted?"
Benchmarking Report - James Multi-Centre Session
Prompt: Identify any teachers who's misfit exceeded 2 standard deviations. In each case provide a report that describes any concerns and recommend appropriate courses of action
    
James Multi-Centre Judge Misfit analysis report
Prompt: "Identify any Items who's misfit exceeded 2 standard deviations. In each case provide a report that describes any concerns and recommend appropriate courses of action. Identify the school that the items relate to also."