MODULE 4 · LESSON 4.6
Turning Triangulated Findings into Recommendations
Estimated time: 25 minutes
Data analysis is useful when it helps people understand a problem and decide what to investigate or do next. Recommendations must be supported by the evidence.
4.6.1 Separate findings from explanations
Suppose a fictional district's reporting completeness falls from 95% to 80%.
- Finding: reporting completeness declined by 15 percentage points.
- Possible explanations: some facilities may have failed to submit reports, or the number of expected reports may have changed.
- What remains unknown: the available figures do not establish why the decline occurred.
- Next step: review facility-level submissions and confirm the expected reporting units for both periods.
This is better than immediately claiming that staff performance has deteriorated.
4.6.2 Build a triangulation matrix
| Question | Facility reports | NGO reports | Community survey | Interpretation |
|---|---|---|---|---|
| Was the service delivered? | Records show service activity | Records show outreach activity | Some respondents report receiving the service | All sources provide relevant but different evidence |
| Are the figures directly comparable? | Counts service contacts | May count outreach contacts | Estimates unique people | Not without further checks |
| What requires investigation? | Possible missing or duplicate entries | Possible overlap with facility records | Sampling and recall limitations | Clarify definitions, coverage and methods |
The matrix does not force the sources to agree. It helps the analyst understand what each source contributes and where additional investigation is needed.
4.6.3 Write an evidence-based recommendation
A weak recommendation says: the district must dismiss the facility staff because reporting completeness declined. That recommendation makes an unsupported assumption about the cause of the decline.
A stronger recommendation says: the district should review facility-level reporting records, confirm the number of reports expected, and identify the facilities responsible for the decline in reporting completeness. Corrective support should be based on the findings of that review. This responds to the observed problem without assuming its cause.
4.6.4 Present findings to decision-makers
- What question was investigated?
- What data sources were used?
- What did the analysis find?
- What limitations affect interpretation?
- What should happen next?
The audience may be a hospital manager, local government health officer, NGO programme director or public health team. Avoid unnecessary technical language. Explain the findings in terms that support action.
This module teaches general public health data triangulation principles. HIMForward Academy is an independent educational platform and does not claim official affiliation with, or endorsement by, the Federal Ministry of Health (FMoH) or the MSDAT initiative.