Chborn in full count 1950 Census

USA
1

Hello,

I am having issues with the variable chborn in the full count 1950 Census. I understand the question was only asked to ever-married sample-line females, age 14+. But even within that subsamples that are too many missing values. For instance if I do “tab chborn if slrec==2 & sex==2 & age>=45 & age<=65 & marst==1” I find that 88% of the responses are n/a. Do you know what could be the reason? Thank you!

Giulia

2

The universe of the variable CHBORN in the 1950 census data is “Ever-married 1-in-6 sample-line (3.30% sample) females age 14+; not available in Alaska or Hawaii.”

From our page on IPUMS USA sample designs:

The 1950 census schedules are contained on 6,278 microfilm reels. Each census page contains information on thirty individuals. Every fifth line on the census page was designated as a sample line, and additional questions for the sample-line individuals on each page appear at the bottom of the form. For the last sample-line individual on each page, there was a block of additional supplemental questions. Thus, 20 percent of individuals were asked a basic set of supplemental questions, and 3.33 percent of individuals were asked a full set of supplemental questions.

Thus, not all ever-married sample-line females age 14+ answered the question about childbearing. Only 1 in 6 sample persons were eligible for this question and many other questions in the 1950 census, and then after determining eligibility based on their location on the census pages, the universe restrictions were applied (age, sex, marital status). We do not currently provide a variable that identifies the 3.33 percent sample-line persons. We will work to make our documentation on this 3.33 percent subsample clearer and more visible, and hope to provide an identifier for the subsample similar to SLREC in the future.

You will need to exclude Alaska and Hawaii from your analysis, since this question was not asked in either state.

3

That makes sense, thank you. I hadn’t realized the full-supplemental subsample was nested within the 20% sample line — that explains the N/A pattern I was seeing.

Two quick follow-ups:

  1. For analyses using chborn on this 3.33% subsample, is SLWT still the appropriate weight, or is there a different weight intended for the full-supplemental questions?
  2. I’m computing children ever born per adult woman, which means adding adult never-married women back in as zeros. Since chborn was only asked of ever-married women in the 3.33% subsample, is there a way to identify the never-married women who belong to that same subsample (e.g., by page/line position), so my denominator matches my numerator? Without a subsample flag I can’t exactly tell which never-married women to include.

Thanks!

4

In the 1950 full count file, you should multiply SLWT by 6 for analysis of variables that were asked of the 1-in-30 (1-in-6 of the original sample persons). This modification ensures you will obtain accurate frequencies for this subpopulation. Other types of estimates (e.g., averages, regression coefficients) won’t be affected by multiplying the weight, so you can use SLWT without modifying it.

We do not currently have a variable or code that allows you to directly identify this subset of sample line persons in the full count 1950 data. We are looking into whether we will be able to provide this at some point. We will also revise our documentation to clarify many of the points you have raised, so thank you for prompting us to review our documentation and variables related to sample line persons in the 1950 data.