Definition
Primary data is information you collect firsthand for a specific M&E purpose, through surveys, interviews, focus groups, or direct observation. Secondary data is existing data collected by someone else for a different purpose that you repurpose for your analysis.
The distinction matters because each type carries different trade-offs. Primary data gives you precise control over what you measure, how you measure it, and the quality standards applied, but it requires significant time and resources. Secondary data can provide immediate insights at minimal cost, but may not align perfectly with your indicators or context.
Why It Matters
Choosing between primary and secondary data is one of the most important cost-benefit decisions in M&E. Misjudging this choice leads to either wasteful data collection when existing data would suffice, or flawed analysis based on data that doesn't answer your evaluation questions.
Secondary data is particularly valuable for baseline studies (historical data can establish pre-intervention conditions), triangulation (independent sources strengthen validity), and rapid assessments where time or budget constraints rule out full primary collection. However, relying exclusively on secondary data risks measuring the wrong things, donor indicators, administrative records, or published statistics rarely match your programme's specific theory of change.
In Practice
Primary data collection includes:
- Programme-specific survey design and implementation
- Key informant interviews with programme beneficiaries and staff
- Focus group discussions on programme experiences
- Direct observation of activities and outputs
- Most significant change stories collected through your monitoring system
Secondary data sources include:
- Government census and household surveys (DHS, LSMS, EMIS)
- Administrative records (health facility reports, school attendance logs)
- Donor databases and sector-wide statistics
- Academic research and evaluation reports from similar programmes
- Satellite imagery and remote sensing data
- Social media and digital platform analytics
Best practice: Use a mixed approach. Secondary data establishes context and benchmarks; primary data captures programme-specific outcomes. For example, a health programme might use national DHS data for baseline health indicators while collecting primary data on programme-specific service quality and patient satisfaction.
Related Topics
- Data Quality Assurance, validity and reliability apply differently to each data type
- Baseline Design, determines whether you need primary collection or can use existing data
- Rapid Assessment, often relies heavily on secondary data
- Data Management, handling and documentation requirements differ by source
Further Reading
- CARE International. (2020). Secondary Data Analysis Guide, Practical guidance on evaluating and using existing data sources.
- World Bank. (2019). Using Administrative Data for M&E, When and how to leverage government and programme records.
- BetterEvaluation. Secondary Data Sources, Comprehensive directory of international data sources by sector.