Three-Day Workshop on Reproducible and AI-aided Health Data Analysis at IQRAA

Background

The IQRAA workshop series was designed to build practical data analysis skills in R for clinicians and health researchers. It runs across multiple days and progresses from data fundamentals through exploratory data analysis to applied statistical methods, using realistic clinical datasets so participants work on problems close to their own.

Approach

The series combines Quarto Reveal.js slide decks, demonstration datasets, hands-on exercises, and a participant handbook. Modules include an AI-basics introduction and a dedicated exploratory data analysis module that walks through the core dplyr verbs (distinct, count, arrange, filter, summarize, group_by, mutate, case_when, joins, and pivot) using the WHO tuberculosis dataset, alongside applied demos such as low-birth-weight analysis and survival analysis.

What we found

• Participants gain fluency in tidyverse data transformation and ggplot2 visualisation.
• Builds an end-to-end EDA workflow from raw data to summary tables and figures.
• Introduces foundational AI concepts relevant to data analysis.

Outputs & impact

• A reusable curriculum: slide decks, demo and raw datasets, exercise scripts with solutions, a data dictionary, figure/PICO reference cards, and a participant handbook (PDF).
• Self-contained HTML decks for re-delivery.
• Public repositories and decks: iqraa-working-with-data-eda (Working with Data: EDA with the tidyverse), iqraa-best-practices-data-science (Best Practices in Data Science), and iqraa-ai-basics (AI Basics).
• Venue: the IQRAA workshop in Kozhikode, Kerala (precise institutional name to be confirmed).