This analysis aims to uncover the most appropriate outdoor PM2.5 exposure data to use in southwestern Oregon to provide accurate estimates of health impacts. The study compares results from analyses using exposure data organized at four different data structures: treatment hospitals, zip codes, airsheds, and smoke dispersion models. Statistical comparisons encompass descriptive statistics, mean differences, correlations, and regression parameters to illuminate the impact of data choices on the health-smoke exposure relationship. The airshed level data outperforms the other datasets in almost all statistical tests. Subject: PM2.5 exposure, wildfire smoke, respiratory health impacts, data comparison
