Data triangulation is when a piece of data or a finding can be verified with several different research methods or data sources, adding credibility to the findings.
- Requires planning, preferably prior to data collection
- Uses different data sources across different levels and time periods
- Linkage over time (panel data) is done to examine program impacts while controlling for individual-level factors at the outset.
- Linkage of couple-level data can be done, although there are both advantages and disadvantages to this approach.
- Linkage of mother and child data is often done; linkage to father’s data is more difficult.
- Survey data sets (for example, household and facility information) can be linked to:
- compare level of health services availability to health outcomes across geographical units
- examine the effects of physical attributes on service utilization
- create time series and panel data to help build causal explanations of program effects
In some cases, it is not appropriate to link data.
- It may not be necessary for some programs in a given context.
- If done with improper methodology or data sources, linking data can provide inaccurate findings.
- Linked analysis is more appropriate for evaluation than for monitoring.