Using the IPUMS Online Data Analysis System
After clicking one of the datasets, specify the following to create a table:
- a row variable
- a column variable (optional): will be cross-tabulated with the row variable
- filters (optional): can be used to include only certain cases in your analysis
- controls (optional): produces a separate table for each category of the specified variable
Examples and screenshots are available in our short demonstration page based on the IPUMS-USA system.
The tabulator can also perform other analyses, such as multivariate regression, correlation matrices, and comparisons of means. See the contextual help menus for more information.
NOTE: The statistics produced by the online analysis system are derived from sample data. They may not match official published statistics.
Datasets available for analysis are listed on the main page for the online system. The first set of links are for individual samples. The links further down on the page apply to datasets that combine multiple samples into one dataset. Tabulations of the country datasets should include the "year" or "sample" variable to yield meaningful results. The regional datasets (e.g., All Africa) combine samples for many countries and are very large. Analyses of the regional datasets should include the "sample" variable. In tabulations of the regional files you can easily exceed the number of cells allowed by the system. You may need to apply filters or recode variables.
When you select a sample, the available variables are accessible in an expandable list on the left side of the screen. Not all variables are available for every sample in the combined datasets. To view full variable documentation, open a second browser instance of the IPUMS variables page, or drill down through the "codebook" link at the top of the screen.
Some samples have important notes about their design or coverage. See the Sample Description page.
By default, all statistics are measured at the person level. Even household variables will yield the number of persons residing in households with the selected characteristics. To get household-level results, use the PERNUM variable as a filter to select only the first person in each household ("pernum(1)") and apply the household weight variable in the Weight field.
Missing values will be excluded from analyses by default. You can choose to include them by checking the appropriate box among the "Output Options" for the statistical procedure.
The IPUMS SDA system allows users to create their own variables using the "Create Variables" tab. These are not automatically removed when new versions of the IPUMS SDA samples are posted; therefore, created variables may be based on IPUMS data that are no longer valid. (The SDA dataset creation date can be found by clicking on the "Codebook" tab, while the date a user-created variable was made can be found in the list of created variables.)
You must be registered for IPUMS-International to use the online analysis system. Apply for access here. When clicking on a sample, you will be prompted to enter your IPUMS username and password. If you have difficulties, your registration may have expired. Confirm that you can log in to the extract system at the "Login" link above.