Accuracy of Commercial Geocoding
Author Information
Author(s): Eric A. Whitsel, Miguel P. Quibrera, Richard L. Smith, Diane J. Catellier, Duanping Liao, Amanda C. Henley, Gerardo Heiss
Primary Institution: University of North Carolina
Hypothesis
How does the accuracy of commercial geocoding vary among different vendors and address characteristics?
Conclusion
Geocoding error depends on the measures used to evaluate it, address characteristics, and vendor selection.
Supporting Evidence
- Vendor A had the highest address match rate at 98%.
- Vendor D had a significantly lower match rate of 30%.
- Mean positional error varied greatly among vendors, with Vendor A showing the lowest mean distance of 1809 meters.
Takeaway
This study looked at how well different companies can match addresses to their correct locations, finding that some do better than others.
Methodology
Addresses were geocoded by four vendors and compared based on match rates, concordance, and distance from established coordinates.
Potential Biases
Potential bias due to the elusiveness of a definitive criterion standard and the possibility of errors in the criterion standards used.
Limitations
The study's criterion standards may have been imperfect, potentially leading to overestimation of match rates and underestimation of mean positional error.
Participant Demographics
Addresses from various sources across 49 U.S. states, including participants from the Women's Health Initiative and the Atherosclerosis Risk in Communities study.
Statistical Information
P-Value
< 10^-4
Confidence Interval
95% CI: 47, 93
Statistical Significance
p<0.0001
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website