Editing ZIP Code (section)

=== Statistics ===
{{As of|October 2019}}, there are 41,702 ZIP Codes in the United States.<ref>{{Cite web|url=https://greatdata.com/product/zip-code-database|title=ZIP Codes|archive-url=https://web.archive.org/web/20191019150049/https://greatdata.com/product/zip-code-database|archive-date=19 October 2019|url-status=dead}}</ref> Due to convenience, ZIP Codes are used not only for tracking of mail, but also commonly for gathering geographical statistics in the United States by some researchers.<ref name="Grubesic1">{{cite journal |last1=Grubesic |first1=Tony H |last2=Matisziw |first2=Timothy C |title=On the use of ZIP codes and ZIP code tabulation areas (ZCTAs) for the spatial analysis of epidemiological data |journal=International Journal of Health Geographics |date=2006 |volume=58 |issue=5 |page=58 |doi=10.1186/1476-072X-5-58|doi-access=free |pmid=17166283 |pmc=1762013 }}</ref><ref name="Chen1">{{cite journal |last1=Xiang |first1=Chen |last2=Ye |first2=Xinyue |last3=Widener |first3=Michael J. |last4=Delmelle |first4=Eric |last5=Kwan |first5=Mei-Po |last6=Shannon |first6=Jerry |last7=Racine |first7=Elizabeth F. |last8=Adams |first8=Aaron |last9=Liang |first9=Lu |last10=Jia |first10=Peng |title=A systematic review of the modifiable areal unit problem (MAUP) in community food environmental research |journal=Urban Informatics |date=27 December 2022 |volume=22 |issue=1 |doi=10.1007/s44212-022-00021-1|doi-access=free |bibcode=2022UrbIn...1...22C }}</ref> ZIP Codes are not created for statistical analysis, and thus their use for statistical analysis is heavily criticized for numerous reasons and advised against as a cartographic practice.<ref name="Chen1"/><ref name="Adams1">{{cite journal |last1=Adams |first1=Aaron |last2=Chen |first2=Xiang |last3=Li |first3=Weidong |last4=Zhang |first4=Chuanrong |title=Normalizing the pandemic: exploring the cartographic issues in state government COVID-19 dashboards |journal=Journal of Maps |date=27 Jul 2023 |volume=19 |issue=3 |pages=1–9 |doi=10.1080/17445647.2023.2235385|doi-access=free |bibcode=2023JMaps..19Q...1A }}</ref> As ZIP Codes are not polygons, but collections of mail routes and points, they are unsuitable for publication or distribution of most data.<ref name="Census Bureau1">{{cite web |title=ZIP Code Tabulation Areas (ZCTAs) |url=https://www.census.gov/programs-surveys/geography/guidance/geo-areas/zctas.html |website=U.S. Census Bureau |access-date=26 August 2023 |archive-date=August 26, 2023 |archive-url=https://web.archive.org/web/20230826203531/https://www.census.gov/programs-surveys/geography/guidance/geo-areas/zctas.html |url-status=live }}</ref> Polygons for ZIP Codes are not released by the USPS and instead interpolated by 3rd party vendors.<ref name="Grubesic1" /> These interpolations introduce topological errors and are not standard between vendors.<ref name="Grubesic1" /> The USPS often discontinues, splits, or otherwise modifies ZIP Codes, making continuous space-time analysis challenging, leading to issues with both the [[modifiable areal unit problem]] (MAUP) and [[modifiable temporal unit problem]] (MTUP).<ref name="Grubesic1" /><ref name="Adams1"/> As the ZIP Codes are postal routing numbers, individuals and organizations without concrete spatial locations may be given a number, making it impossible to associate demographic data with them.<ref name="Grubesic1" /> Demographic data is inconsistent between ZIP Codes, and no effort is made to ensure they are proper enumeration units for analysis.<ref name="Grubesic1" /> As ZIP Codes are not made with the same considerations as other enumeration units, and is not possible without committing the [[ecological fallacy]].<ref name="Grubesic1" /><ref name="Chen1"/> This again becomes an issue with the MAUP. They have been found not to have significant correlations with health indicators, which can lead to poor conclusions.<ref name="Chen1"/> Despite these issues, ZIP Codes remain popular among researchers in fields such as public health due to their convenience, public familiarity with them, ability to anonymize subject addresses through aggregation, and possible ignorance of more appropriate enumeration units on the part of researchers.<ref name="Grubesic1" /><ref name="Adams1" />

In an attempt to satisfy demand "by data users for statistical data by ZIP Code area", the [[United States Census Bureau|U.S. Census Bureau]] calculates approximate boundaries of ZIP Code areas, which it calls [[ZIP Code Tabulation Area]]s (ZCTAs).<ref name="Census Bureau1" /><ref name="Grubesic1" /> Statistical census data is then provided for these approximate areas. The geographic data provided for these areas includes the [[latitude]] and [[longitude]] of the center-point of the ZCTAs. ZIP Codes are inherently discrete or ''point-based'' data, as they are assigned only at the point of delivery, not for the spaces between the delivery points. The United States Census Bureau then interpolates this discrete data set to create polygons by attempting to match ZIP Code extents with Census blocks.<ref name="Grubesic1" /> The resulting aereal units represent the approximate extent of the ZIP Code, which are combined to use for mapping and data presentation. The process of creating ZCTAs and their use for statistical analysis is heavily criticized in the literature. First, the creation of ZCTAs from Census blocks encounters issues when a Census block straddles multiple ZIP Codes.<ref name="Grubesic1" /> Addressing this is another instance of the MAUP, and the solution of dividing aggregate units between ZIP Codes causes some individuals to fall into ZCTAs that do not match their ZIP Code.<ref name="Grubesic1" /> The creation of these units is therefore committing the [[ecological fallacy]] by attempting to disaggregate aggregate data. As ZIP Codes are not continuous, not everyone in the United States has one; there are ZIP Codes for non-populated or geographic areas, resulting in there not being one ZCTA for every ZIP Code.<ref name="Grubesic1" /> ZCTAs are not updated as frequently as the USPS updates ZIP Codes, resulting in further temporal analysis issues when ZIP Codes change during a study period.<ref name="Grubesic1" /> Datasets providing a similar approximate geographic extent to ZCTA are commercially available. Despite these issues, ZCTAs are still very popular with researchers in fields like epidemiology, and among government agencies, with some states employing them to publish and distribute public health data during the COVID-19 pandemic.<ref name="Adams1" />