Start by defining the research question and identifying the topic, geographic level, and timeframe that align with your study. NaNDA datasets cover a wide array of neighborhood characteristics, including demographics, socioeconomic factors, built environment measures, and neighborhood-level amenities like parks and public transportation.

1. Research Question: What aspect of neighborhoods are you studying? Whether it’s social factors, environmental characteristics, or community resources, understanding the focus of your research will help narrow down your options.
2. Geographic Level: NaNDA datasets are available at different geographic levels (e.g., census tracts, ZIP Code Tabulation Areas, and sometimes school districts or counties). Consider which level of detail is most relevant to your study.
3. Timeframe: NaNDA datasets span different years. Be sure to choose a dataset that fits the time period relevant to your research.

Browse our data catalog to explore detailed descriptions of each dataset. These will help clarify the topics, geographic levels, and timeframes covered, ensuring you find the dataset that best fits your needs.

NaNDA datasets are typically available at the census tract or ZIP Code Tabulation Area (ZCTA) level, though some may also include other geographic units like school districts or counties. When using NaNDA with individual-level source data, select the geographic level that corresponds to the Geographic Identifiers (GEOIDs, e.g., census tract FIPS codes) in the source data.

If multiple GEOIDs are present in the source data (e.g., both census tract FIPS codes and ZCTAs), the appropriate choice depends on your research question. Census tracts are smaller than ZCTAs and capture more localized conditions. For example, when studying neighborhood amenities like grocery stores or coffee shops, the census tract may be the most relevant geographic level. Some resources and processes, however, function at a broader spatial scale, such as ZCTAs or even counties. For example, healthcare access might be better captured at the ZCTA level, as people often travel beyond their immediate neighborhood to visit a doctor’s office.

To learn more, check out our blog post: Understanding Geographies in NaNDA: Tract10, Tract20, ZCTA10, and ZCTA20

NaNDA datasets use geographic identifiers (GEOIDs) as the unique identifier for each entry. Most often, NaNDA data includes either a census tract FIPS code or a ZIP Code Tabulation Area (ZCTA), but in some cases, datasets may use school districts, counties, or other geographic identifiers.

Data organized by the same unique identifier will merge seamlessly with NaNDA using statistical software or data processing tools. If your data doesn’t include these identifiers, you may need to geocode your observations to assign the correct geographic identifier before merging.

Missing data in NaNDA often stems from two main issues:

• Zero or Missing Land Area: In some census tracts—especially in U.S. territories like Puerto Rico or the Virgin Islands, or along shorelines and water bodies—the land area is either zero or missing. This affects variables like area density because without a defined land area, the density calculation can’t be completed.
• Zero Population Tracts: Similarly, some density variables are set to missing when the tract population is zero.

The same rules apply across various variables in NaNDA. To handle missing data, start by checking the geographic context of your data, and then consider applying statistical techniques like imputation to mitigate its impact on your analysis. Learn more about this topic on our blog post: Understanding Missing Data in NaNDA: A Closer Look at Area Density and Other Variables.

Got it? Missing data might be a puzzle, but understanding its pieces makes for a clearer picture!

Yes! NaNDA datasets are available in tabular formats (e.g., CSV, Stata, and others), with geocoding already handled. GIS software is not necessary for analyzing NaNDA data. Once merged with other datasets and analyzed, users can create maps using GIS software like ArcGIS or QGIS to visually enhance results.

To stay informed about updates to datasets, subscribe to NaNDA News Updates. We send notifications when new data are published, as well as other relevant updates and announcements. While updates vary depending on the dataset, some are updated annually, while others may be updated more or less frequently. Subscribing ensures you’ll receive timely information on data availability and updates as soon as they’re published.

Each NaNDA dataset includes a digital object identifier (DOI), which should be included in the citation. Citation management tools like Zotero and EndNote automatically generate citations in the required style when provided with a DOI. In addition to the DOI, citations generally should include the dataset title, authors, publication year, and the repository (ICPSR). Always check the specific citation instructions for each dataset.

While it’s important to cite the specific dataset used, if citing NaNDA as a whole is absolutely necessary (e.g., in grant proposals with page limits), the general NaNDA DOI can be used: http://doi.org/10.17616/R31NJNQA. However, citing specific datasets is the preferred and most accurate approach.

Users agree to ICPSR’s general terms of use upon data download.

For ICPSR-related questions (e.g., issues with downloading datasets, accessing files, or technical problems with the ICPSR site), their support team is available Monday to Friday, 8:00 a.m. to 5:00 p.m. ET. Contact them at [email protected] or call 734-647-2200. ICPSR aims to respond to inquiries on the same day.

For NaNDA-specific questions, follow these steps:

1. Read the documentation PDF that accompanies the dataset downloaded—it often contains answers to common questions.
2. If the documentation doesn’t resolve the issue, reach out to the NaNDA team at [email protected]. The goal is to respond within 24 business hours.
3. For further assistance, sign up for office hours.