Situation: A public health nurse at the Rural Health Unit (RHU) is preparing a study of handwashing and diarrhea among children under 5 years in a municipality of 18 barangays. She has a complete list of households in each barangay. A provincial report shows that barangays with a higher proportion of households with piped water had lower rates of childhood diarrhea. A barangay health worker (BHW) concludes that any child in a household with piped water has a lower risk than a child in a household without it. What is the MAIN flaw in this conclusion?

The report compares barangays, not individuals: it is an ecological analysis using group-level data. Concluding that the same relationship holds for each household commits the ecological fallacy, because the children who got diarrhea in a barangay may not be the ones living without piped water. Individual-level data (for example, a cohort or case-control study of households) are needed for that conclusion.

Core concept: ecological fallacy
The provincial report describes a relationship at the barangay level: barangays with more piped-water households tended to have lower diarrhea rates. The BHW then jumps from that group-level pattern to a claim about each individual child. That jump is the flaw.

When correlation or regression values are computed from group averages, they can differ from the values computed using individual scores. ^[1][2] This is the formal definition of the ecological fallacy: assuming that a relationship observed at the aggregate level automatically holds at the individual level.

In this barangay example, the children who actually developed diarrhea in a barangay may not be the same children living in households without piped water. A barangay could have a high proportion of piped-water households overall, yet the diarrhea cases could still cluster among the few households lacking piped water—or among piped-water households with poor storage or hand hygiene. The group-level data cannot answer that question.

Key point! The report compares barangays, not individual households. To conclude that a specific child in a piped-water household has lower risk, the nurse would need individual-level data—for example, a cohort or case-control study that links each child’s diarrhea outcome to that child’s own water source and other exposures.

Why the other options are not the main flaw

Option	Why it is not the main flaw
1. Recall bias inflated diarrhea reports in poor barangays	This is a possible measurement problem, but the scenario does not describe how diarrhea was reported or whether recall differed by barangay wealth. The stated error is about the level of inference, not data collection bias.
3. Rates were computed with the wrong population base	The report compares barangay-level rates, which is a legitimate ecological comparison. The problem is not the denominator but the inference from group to individual.
4. Barangays with piped water underreported their cases	Again, this is a reporting or surveillance concern that is not supported by the scenario. The conclusion would still be flawed even if reporting were perfect, because the level of analysis is wrong.

Why this matters for nursing and public health practice
In community health, nurses frequently see aggregate reports—barangay profiles, district health statistics, or WASH coverage maps. These are useful for planning and targeting interventions, but they cannot establish individual risk. For example, a meta-analysis of childhood diarrhea in Ethiopia identified multiple individual-level risk factors, including water source, sanitation, and hygiene practices, and emphasized that safe piped water is not universally available in developing settings. The association between WASH factors and diarrhea must be examined at the household or child level to guide individual counseling.

The sampling distribution of a correlation coefficient computed from group averages can differ from that computed from individual scores, which is why the ecological fallacy persists even when a sample is drawn from the same population. ^[1][2] In other words, the barangay-level correlation and the household-level correlation are not interchangeable.

Watch out! In licensure exams, any scenario that moves from a group statistic to a statement about an individual within that group is testing the ecological fallacy. The correct response is to identify the mismatch between the unit of analysis (barangay) and the unit of inference (individual child).