Householder education

Hi I am trying to look at cbhhtype crosstabbed with education. However I want the education of the householder, not all the people living in the specific type of household noted as cbhhtype. Is this possible?

To determine the educational attainment of the householder by household type as defined by the Census Bureau, you can tabulate CBHHTYPE and EDUC for only those respondents identified as the householder in the variable RELATE. Respondents with a RELATE value of 1 are the head of household/householder.

The default data structure for IPUMS USA files is a file that is rectangularized on the person. Assuming you choose this structure, the household-level variables (including CBHHTYPE) will be attached to the person record.

You should use person weights (PERWT) with person-level analyses, such as calculating the share of householders with different educational attainments/household types. You should use household weights (HHWT) with household-level analyses, such as calculating the share of households by type with different householder educational attainment characteristics. You should use replicate weights to calculate standard errors or confidence intervals from ACS data.

Thank you–Would I do that here: I am unsure of which website to use to extract these data–thank you.

The link you provided is to the online data analysis tool (SDA), which you can use to analyze data online. Alternatively, you can work with IPUMS data by creating, downloading, and analyzing a microdata extract. The online data analysis tool is a quick, easy way to create basic tables and graphs using IPUMS USA data, and is appropriate for many analyses. Users wishing to conduct more complex analyses, such as hypothesis testing, should create a custom microdata extract and analyze it using a statistical software package such as R, Stata, SPSS, or SAS.

You can get started learning how to use the online data analysis system by reading the instructions or watching a short video tutorial. If you prefer to create and analyze microdata in extract form in a stats package, you can start with this video on how to use the data extract system and this blog post on opening your data file. The analysis you described in your original forum post is feasible using either the online data analysis system or by creating and analyzing an extract.

Thank you. My final question, is how do I get data on Monterey County (County ICPSR=0530) in California (State ICPSR=71)? When I filter on that county value there is no data–do I have to aggregate two different areas, if so which ones? Thank you.

And lastly, why is it that the SAMPLE variable shows 0 records for ACS5 year estimates?

I will first answer your question about Monterey County, California. County is not explicitly identified in the public use microdata from 1950 onward. PUMA (Public Use Microdata Area) is the smallest geographic area identified from 1950 onward. IPUMS geographers can sometimes infer other geographic areas, such as county, based on PUMA. However, this identification depends on how counties overlap with PUMAs (in the case of counties, we only identify them if there is a perfect alignment between PUMA(s) and a county). Note that PUMA definitions are updated by the Census Bureau following each decennial census.You can see a list of identified counties by IPUMS USA sample here. Monterey County is not identified in any ACS samples. You can read more about identification and definition of counties in the COUNTYICP metadata.

If you are specifically interested in Monterey County and do not need to use person-level data, you may consider accessing data through IPUMS NHGIS. IPUMS NHGIS provides summary data from the U.S. census and American Community Survey aggregated at various geographic levels, including county. Data from IPUMS USA are microdata, meaning each line or observation is a person or household. IPUMS NHGIS data are summary data, and each line or observation is a specific geographic area (such as Monterey County or the state of Texas). From IPUMS NHGIS, you can access tables that display information about each geographic area, such as its population, or the number of people living in the county with different educational attainment levels. IPUMS NHIGS data do not allow you to identify characteristics of individuals (and therefore determine who the head of household is), but the data are more geographically precise.

I’m not sure I understand your second question about SAMPLE. If you have restricted your extract or SDA results to show only Monterey County California, I would expect all observations to be zero in years where the county is not identified. Could you provide some more information about the estimate you’ve created (are you using the SDA or an extract, what sample(s) are you looking at, etc.)?