Lecture 3.1.12: Partner Scoping & Secondary Data Discovery
Why It Matters
Knowing which secondary health datasets already exist prevents costly duplicate surveys and enables faster, evidence‑based decision‑making for policymakers and researchers.
Key Takeaways
- •Check existing data before launching costly primary surveys.
- •Partner scoping maps data holders and prevents duplication of effort.
- •DHS, MICS, and DHIS2 are the three core secondary sources.
- •Document findings in a simple Excel inventory for transparency.
- •Triangulate across all three sources to mitigate biases and gaps.
Summary
Partner scoping and secondary data discovery are presented as the first step in health data science, emphasizing that researchers should ask whether needed data already exist before launching expensive primary surveys. The lecture outlines the three cornerstone repositories for low‑ and middle‑income countries—Demographic and Health Surveys (DHS), Multiple Indicator Cluster Surveys (MICS) and District Health Information System 2 (DHIS2)—and explains how each serves a distinct purpose.
The speaker argues that secondary data save time and money, citing free, nationally‑representative household surveys covering tens of thousands of respondents. Types of secondary data include household surveys, routine health system records, civil registrations, and research repositories. A step‑by‑step partner‑scoping workflow—define the question, map the ecosystem, search known repositories, contact data custodians, and build a data inventory—ensures systematic coverage of existing sources.
Illustrative analogies compare secondary data to buying vegetables at a market and DHIS2 to a restaurant order book, highlighting that each source captures only part of the picture. Real‑world examples reference Pakistan’s DHS stunting data, Afghanistan’s reliance on MICS, and the need to avoid ecological fallacy, denominator errors, and outdated information. The lecture stresses ethical considerations, data‑sharing agreements, and the importance of understanding original consent.
The practical outcome is a structured data inventory and gap analysis that can answer up to 80% of research questions without new surveys. By triangulating DHS, MICS, and DHIS2, analysts mitigate individual biases and gain a fuller view of health indicators, while respecting access rules and ethical constraints. Mastery of data discovery, not just analysis, is positioned as the most underrated skill for health data scientists.
Comments
Want to join the conversation?
Loading comments...