Positional accuracy of geocoding from residential postal codes versus full street addresses

Area of Science:

• Geographic Information Systems (GIS)
• Spatial Analysis
• Public Health Research

Background:

• Postal codes are frequently the sole geographic data for linking populations to environmental or contextual information.
• Assessing factors influencing postal code spatial assignment accuracy is crucial for research.

Purpose of the Study:

• To evaluate the impact of delivery mode, representative point type, and community size on postal code geocoding accuracy.
• To determine how these factors affect the precision of spatial assignment for Canadian census data.

Main Methods:

• Utilized Postal Code Conversion File Plus (PCCF+) for 2011 Canadian Census data.
• Sampled 1,004 individuals, ensuring at least 90 observations per category for delivery mode, representative point type, and community size.
• Calculated geocoding positional accuracy by comparing PCCF+ assigned locations with full street addresses (Google Maps) using GIS.

Main Results:

• Positional accuracy was predominantly influenced by the mode of postal delivery.
• Rural and mixed delivery modes exhibited substantially higher geocoding errors compared to urban modes.
• Low accuracy was observed in rural/small-town Canada and for dissemination area centroids, linked to rural/mixed delivery modes.

Conclusions:

• Postal code geocoding accuracy is variable and influenced by delivery mode.
• Geocoding imprecision can lead to misclassification, contingent on the spatial resolution of contextual data.
• Study design should consider required spatial resolution to identify subpopulations with inadequate positional accuracy.