Date of Award




Document Type

Master's Thesis

Degree Name

Master of Science (MS)


Department of Epidemiology and Biostatistics



Content Description

1 online resource (v, 47 pages)

Dissertation/Thesis Chair

Louise Anne McNutt

Committee Members

Bruce Coles, Ping Du, Millicent Eidson


ABSMs, Geographic Level, Gonorrhea, Methodology, Medical geography

Subject Categories



This study compared the association between residential area factors and gonorrhea, aggregated at both the census tract versus ZIP code levels. Prior research has found that census block groupings are most optimal in neighborhood-level studies, followed by census tracts and ZIP codes. Because several healthcare databases currently only provide ZIP code level data, the aim of this study was to examine whether there exists a meaningful difference between the interpretation of census tract and ZIP code aggregate data. Our study measured the association between socioeconomic factors and gonorrhea in New York State, exclusive of New York City (NYC), census tracts (N=2660) and ZIP codes (N=1416), from 1999 to 2001. Both analyses were stratified by level of urbanicity, based on the guidelines of the United States Department of Agriculture's (USDA's) Rural-Urban Continuum. Overall, we found few differences in results based on the aggregation levels. Similar to previous studies, percent below poverty and percent black were the most consistent predictors of gonorrhea in a geographic region defined by either census tract or zip code. However, estimates for these variables in the rural stratification were unstable for both geographic levels, but notably more pronounced within the ZIP code analysis. This indicated likely underestimation of the association between SES/race and gonorrhea in the rural regions. It is not clear which level of aggregation works best for these rural neighborhoods. As the amount of data available to researchers expands in the upcoming years, especially following the Census 2010, it will be important to recognize and adjust for potential methodological issues such as the ones in this study.

Included in

Epidemiology Commons