Nutrition and Obesity in Latin America
Project Description
This project uses real public health data to investigate the rise of obesity across 20 Latin American countries. Learners pull data directly from the World Health Organization’s Global Health Observatory, clean and reshape it, and build a series of charts that answer concrete questions: How have obesity rates changed over time? Which countries carry the heaviest burden? Are women and men affected equally? And does the same country struggle with child undernutrition and adult obesity at once — the “dual burden” of malnutrition?
The work moves through a full data science pipeline, from collecting raw API data to producing a final data story. Along the way learners write code to fix messy real-world data, interpret what the charts actually mean, and finish by zooming in on a single country to make a policy recommendation grounded in the numbers. No prior public health knowledge is needed — the project is designed to be approachable while still touching every core step of a real analysis.
Dataset
The data comes from the WHO Global Health Observatory (GHO), accessed live through its free OData API (no login or download required). The project draws on six health indicators across 20 Latin American countries, spanning several decades:
- Adult obesity prevalence (BMI ≥ 30)
- Adult overweight prevalence (BMI ≥ 25)
- Child and adolescent obesity (ages 5–19)
- Mean adult BMI
- Child stunting prevalence
- Adult underweight prevalence
Because the data is fetched directly from the API each time the notebook runs, learners work with the same source health researchers and policymakers rely on, rather than a pre-packaged file.
Relevant Skills You May Apply
Basic Python, familiarity with running code in a notebook, and comfort reading short blocks of code. Any prior exposure to pandas, charts, or working with tables will help but is not required — the notebook walks through each step.
Skills You May Gain
- Collecting data from a live web API and combining multiple sources into one dataset
- Preprocessing: selecting, renaming, and retyping columns; reshaping data into an analysis-ready form
- Data cleaning: handling duplicates, missing values, and impossible values
- Exploratory data analysis: summary statistics, data-availability checks, and reading what the data can and cannot tell you
- Data visualization with matplotlib and seaborn: line charts, bar and grouped bar charts, scatter plots, and heatmaps
- Interpreting results and translating findings into a clear data story and a policy recommendation
Total Time
Roughly 4 to 6 hours. The project can be completed in one sitting or split across sessions, and several sections are well suited to working through as a team.
Milestones
- Data Collection — Import libraries and fetch the six indicators from the WHO API for all 20 countries.
- Data Preprocessing — Select and rename columns, fix the child obesity age groups, and isolate the both-sexes records for the main analysis.
- Data Cleaning — Remove duplicates, drop missing values, and check for out-of-range values before analyzing.
- Exploratory Data Analysis — Inspect the data’s shape and range, and use a heatmap to check how complete the data is across countries and indicators.
- Data Visualization — Build six charts telling different parts of the obesity story, complete the guided fill-in exercises, write a one-country data story card, and (optionally) turn it into a presentation poster.
Deliverables
- A completed Colab notebook with all code cells run, every
### YOUR CODE HERE ###and### TODO ###filled in, and the reflection questions answered.
- A completed data story card focused on one chosen country, including three key findings and a policy recommendation.
- Optional: a visual poster summarizing the chosen country’s story for a general audience.
Before editing anything, all learners must open the notebook and select “File” → “Save a copy in Drive” so they are working on their own personal copy in their Google Drive. Each member of a team should make and edit their own copy rather than working in the shared original. Once finished, completed materials are submitted to the NSDC HQ team to receive a virtual certificate of completion.