What is it about?
We visited 425 homes in Nadowli-Kaleo District, in Ghana's Upper West Region, and asked people to rate their drinking water from "very poor" to "very good." During the same visit, a trained field worker checked the water for visible warning signs: cloudiness, colour, smell and floating debris. The field worker didn't know how the household had rated its water. Then we compared the two. Overall, people's ratings moved in the right direction; water with more warning signs got lower ratings. But household by household, only about half matched what the checklist showed, and most of the mismatches came from people rating their water worse than the checklist suggested. Schooling, income and years lived in the community made no reliable difference to accuracy. One caveat: the checklist flags risk. It isn't a lab test for germs.
Featured Image
Photo by Lisa Marie Theck on Unsplash
Why is it important?
Health programmes often treat what people say about their water as a stand-in for how safe it is. Our results suggest that's a shaky shortcut. People's judgements get the big picture right, but they can't reliably tell you which particular household is at risk. We also found that education level doesn't predict who judges well, so it's a poor way to decide who needs support. What's more likely to help is simple, local water testing with the results shared back to households, so they have something to go on besides sight, smell and taste. Very few studies in West Africa have measured this gap between perception and observed risk directly. It matters for Ghana's progress toward SDG 6, clean and safe water for everyone.
Perspectives
I went into this study half-expecting education to make the difference. I thought people with more schooling would read their water more accurately. The data didn't back that up, and honestly, it changed how I think about water-safety messaging. Almost everyone we spoke to already suspected something was wrong with their water. Concern wasn't what they lacked. A way to check was. I'd love to see community water testing become normal in places like Nadowli-Kaleo, and I hope this paper pushes programme designers a little in that direction. I'm grateful to the households who welcomed us in, and to the enumerators who did the fieldwork, often in tough conditions.
Abdullah Alhassan
University of Business and Integrated Development Studies, Bamahu-Wa, Ghana
Read the Original
This page is a summary of: Perceived drinking-water safety and observed contamination risk: Evidence of an imperfect calibration gradient in rural Ghana, PLOS Water, September 2026, PLOS,
DOI: 10.1371/journal.pwat.0000590.
You can read the full text:
Resources
The dataset (S1 Data)
De-identified processed analytic dataset used to reproduce reported tables and models.
S2 File.
Python analysis script reproducing all reported descriptive statistics, correlation results, regression models, and figures.
S3 File.
R analysis script reproducing all reported descriptive statistics, correlation results, regression models, and figures.
Contributors
The following have contributed to this page
