art photo



I am using the IPUMS data and for 1970 the only file available is the 1970 metro, form 2.

Is this a 1% file?


First, when you open the interface to create an extract in IPUMS, the default for Sample Selection is "the most commonly requested samples." For 1970, the default choice is 1970 Form2 metro.

If you are interested in a national analysis and are finding all the variables you need, this will be fine. If you need to use states as your unit of analysis and/or are finding that some items are not available, go back and change the button on the Sample Selection List to "all of the samples."

Second, a bit of background on the 1970 samples:

The 1970 sample had 6 different samples:

State         Form1  Form2  
Metro         Form1  Form2  
Neighborhood  Form1  Form2  

Some items - race, sex, age - were on all of the questionnaires. However, other items are specific to one of the questionnaires (thus, form1 or form2). Depending on which form you want, you'll select that sample. On occasion, you'll find that you can't do certain cross tabs as the items are questionnaire-specific. For instance, mother's birthplace, father's birthplace and mother tongue can only be found on the Form 2 file (15% sample), but citizenship and year of immigration are only available from the Form 1 data (5% sample).

Take a look at the variables to determine which sample you'll want to use:

State allows you to look at the data across all 50 states. Metro is a county group file, which allows you to look at groupings of counties so that you can study something like Appalachia, or to examine different parts of the state of Texas to each other, or to examine specific metropolitan areas. The neighborhood sample includes characteristics of the neighborhood on the person record. It has never been used much.

Form 1 is also known as the 5% sample and Form 2 is also known as the 15% sample. In the original data collection, Form 1 was distributed to 5% of the households and Form 2 to 15% of the households. However, for the public use files, the sample sizes are the same. Both are 1% files.

Take a look at the sample descriptions for more details:

If you wanted to look at very simple items, you could pull off the items across all 6 samples and end up with a 6% sample. Right now, you have what is the equivalent of a 1% sample.

Annotated Resources:

Direct Links:

Related Question Groups: