Estimating the effectiveness of health-risk communications with propensity-score matching: application to arsenic groundwater contamination in four US locations

J Environ Public Health. 2014;2014:783902. doi: 10.1155/2014/783902. Epub 2014 Sep 30.


This paper provides a demonstration of propensity-score matching estimation methods to evaluate the effectiveness of health-risk communication efforts. This study develops a two-stage regression model to investigate household and respondent characteristics as they contribute to aversion behavior to reduce exposure to arsenic-contaminated groundwater. The aversion activity under study is a household-level point-of-use filtration device. Since the acquisition of arsenic contamination information and the engagement in an aversion activity may be codetermined, a two-stage propensity-score model is developed. In the first stage, the propensity for households to acquire arsenic contamination information is estimated. Then, the propensity scores are used to weight observations in a probit regression on the decision to avert the arsenic-related health risk. Of four potential sources of information, utility, media, friend, or others, information received from a friend appears to be the source of information most associated with aversion behavior. Other statistically significant covariates in the household's decision to avert contamination include reported household income, the presence of children in household, and region-level indicator variables. These findings are primarily illustrative and demonstrate the usefulness of propensity-score methods to estimate health-risk communication effectiveness. They may also be suggestive of areas for future research.

MeSH terms

  • Arsenic / analysis*
  • Consumer Behavior
  • Environmental Exposure*
  • Filtration
  • Groundwater / chemistry*
  • Health Communication / standards*
  • Health Knowledge, Attitudes, Practice
  • Humans
  • Models, Theoretical
  • Propensity Score
  • Risk Reduction Behavior*
  • United States
  • Water Pollutants, Chemical / analysis*


  • Water Pollutants, Chemical
  • Arsenic