Public use microdata is a set of raw data that have not yet been tabulated into a pre-made table. It is when the Census Bureau takes responses received from surveys, removes personal information so you cannot identify any specific person/household, and puts additional edits in place to protect the confidentiality of respondents. Then the Census Bureau provides the set of raw data for you to conduct your own statistical analysis. Working with microdata requires you to select variables, create categories (recode variables), determine output format, and apply statistical weights.

Before using microdata, here are a few tips:

• Use pre-made statistics through the aggregate data API or pre-made tables on data.census.gov whenever possible: The data will be more accurate in these statistics aggregated by the Census Bureau compared to a custom table you make. The process to find pre-made tables and statistics is also easier than creating custom statistics through public use microdata.
• Geographic coverage in microdata is limited: You can access data for the nation, region, division, and states. A few surveys provide data for public use microdata areas (PUMAs) and selected counties.