Perceived drinking-water safety and observed contamination risk: Evidence of an imperfect calibration gradient in rural Ghana
- 1Department of Environment and Resource Studies, Faculty of Integrated Development Studies, University of Business and Integrated Development Studies (UBIDS), XW-1147-8901, Bamahu-Wa, Ghana
Abstract
In low-resource settings, drinking-water safety decisions are frequently shaped by subjective perceptions rather than objective contamination data, yet this correspondence is poorly characterized. This cross-sectional study (n = 425) in Nadowli-Kaleo District, Upper West Region, Ghana examined concordance between perceived drinking-water quality and field-observed sanitary contamination risk, together with socioeconomic determinants of risk appraisal calibration. Perceived quality was rated on a 1–5 Likert scale; field-observed sanitary contamination risk was aggregated into a 0–4 ordinal risk index from a structured observation checklist (a proxy for, not a direct microbiological measure of, water quality), reverse-scaled, with perceptions classified as calibrated when |gap| ≤ 0.5. Perceived quality was strongly and inversely associated with observed contamination risk (Spearman’s ρ = −0.780; p < 0.001), indicating that household perceptions tracked broad contamination-risk gradients well. However, agreement at the individual household level was weaker: only 49.4% of households met the calibration criterion (mean gap = −0.86) under our primary threshold, and this proportion was sensitive to the threshold chosen (ranging from 26.1% at ±0.3 to 63.1% at ±1.0). Educational attainment showed no reliable association with calibration accuracy (no formal education 55.5%; secondary 44.1%; postgraduate 12.5%, n = 8; Pearson χ²(4) = 7.24, p = 0.124), and a quadratic logistic model provided no evidence of a non-linear (inverted-U) relationship (quadratic OR = 0.94; p = 0.462); income and residence duration were likewise not significant predictors. These findings suggest that perception-based metrics alone may be insufficient for individual-level risk assessment, and that common socioeconomic proxies do not reliably identify well-calibrated households, calling for objective monitoring and tailored risk communication to advance SDG 6.
Related
- Drinking waters contamination by per‐ and poly‐fluoroalkyl substances of the anthropized area of Fos‐Berre, France
- A Weighted Estimation Approach to Correct for Imperfect Mutation Detection Due to Intra-tumor Heterogeneity in Tumor Sequencing
- Neurostructural correlates of real-life social drinking risk in alcohol use disorder
- Neurobiological alterations and autism-like phenotype in mice following regulatory-limit exposure to fungicides in drinking water
- Growth, Heavy Metal Accumulation, and Translocation in Chamomile (Matricaria recutita L.) and Safflower (Carthamus tinctorius L.) Under Severe Cd and Pb Contamination in an Alkaline Soil
BibTeX
@article{alhassan2026perceived,
title = {{Perceived drinking-water safety and observed contamination risk: Evidence of an imperfect calibration gradient in rural Ghana}},
author = {Abdullah Alhassan and Ziemah Boyeh and Abdul-Kadri Yahaya},
journal = {PLOS Water},
year = {2026},
month = sep,
doi = {10.1371/journal.pwat.0000590},
url = {https://journals.plos.org/water/article?id=10.1371/journal.pwat.0000590},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Perceived drinking-water safety and observed contamination risk: Evidence of an imperfect calibration gradient in rural Ghana
AU - Abdullah Alhassan
AU - Ziemah Boyeh
AU - Abdul-Kadri Yahaya
PY - 2026
DA - 2026/09/08
JO - PLOS Water
DO - 10.1371/journal.pwat.0000590
UR - https://journals.plos.org/water/article?id=10.1371/journal.pwat.0000590
AB - In low-resource settings, drinking-water safety decisions are frequently shaped by subjective perceptions rather than objective contamination data, yet this correspondence is poorly characterized. This cross-sectional study (n = 425) in Nadowli-Kaleo District, Upper West Region, Ghana examined concordance between perceived drinking-water quality and field-observed sanitary contamination risk, together with socioeconomic determinants of risk appraisal calibration. Perceived quality was rated on a 1–5 Likert scale; field-observed sanitary contamination risk was aggregated into a 0–4 ordinal risk index from a structured observation checklist (a proxy for, not a direct microbiological measure of, water quality), reverse-scaled, with perceptions classified as calibrated when |gap| ≤ 0.5. Perceived quality was strongly and inversely associated with observed contamination risk (Spearman’s ρ = −0.780; p < 0.001), indicating that household perceptions tracked broad contamination-risk gradients well. However, agreement at the individual household level was weaker: only 49.4% of households met the calibration criterion (mean gap = −0.86) under our primary threshold, and this proportion was sensitive to the threshold chosen (ranging from 26.1% at ±0.3 to 63.1% at ±1.0). Educational attainment showed no reliable association with calibration accuracy (no formal education 55.5%; secondary 44.1%; postgraduate 12.5%, n = 8; Pearson χ²(4) = 7.24, p = 0.124), and a quadratic logistic model provided no evidence of a non-linear (inverted-U) relationship (quadratic OR = 0.94; p = 0.462); income and residence duration were likewise not significant predictors. These findings suggest that perception-based metrics alone may be insufficient for individual-level risk assessment, and that common socioeconomic proxies do not reliably identify well-calibrated households, calling for objective monitoring and tailored risk communication to advance SDG 6.
SN - 2767-3219
ER -