I am all new to this and have been trying to extract correct data for 2 hole days now, hope someone can help with the variables.
I need to extract the total sample size of mothers to a 4-year old children for all states from 1980-2007. I cannot get the correct total sample size, which I have been informed is that the total sample size of mothers in that period is 79.917, but that is not what I get when trying. The information I got is that Personal sample weights are used in the data aggregation.
I also need to know labor force participation rate, hours worked, employment rate, family total income (in real term), mothers age and education level, race, food stamp recipients
I really hope someone can help with the variables so I can get the correct data
Thanks in advance
I assume you are using data from the ASEC. What is your source for the sample size of mothers? And what sample size are you finding?
You can identify mothers of four-year-old children using a combination of AGE, MOMLOC, and PERNUM. Exactly how to do this depends on the software you are using, but there is some example code to do a similar linkage near the end of this page. That page is about IPUMS USA, but the process is essentially the same for IPUMS CPS.
For the other variables you are interested in, you should look at the following variables:
LABFORCE for labor force participation rate
EMPSTAT for employment rate
For hours worked, you can use WKSWORK1 for weeks worked last year, and UHRSWORKLY for usual hours worked per week last year. You can also use UHRSWORKT for the current usual hours worked (this will refer to the survey date in March, instead of the previous year).
FTOTVAL for family total income
AGE for age
EDUC for education level
RACE for race
FOODSTMP for food stamp recipiency
Hi Matthew, I wanted to give sample R code for identifying mothers of children under 14 y/o. Please let me know if this looks errored at all!!
It seems to me that the R code given at the bottom of the page you mention (IPUMS USA) for connecting parents with children is a bit flawed. While I don’t read the other languages offered, I was able to come up with this R alternative:
This code chunk makes a df that takes the characteristics of children and renames PERNUM to be able to be merged back onto mothers of own children living in their same HH.
temp ← data %>%
select(YEAR, SERIAL, MOMLOC, AGE) %>%
rename(PERNUM = MOMLOC) %>%
filter(AGE < 14 & PERNUM > 0) %>% # captures children under 14 y/o who live with their mothers
group_by(YEAR, SERIAL, PERNUM) %>%
summarize(own_kids_u14 = n(), .groups = “drop”)
This code chunk joins the number of children back on to the original df and connects the number of children with their mother.
data_moth ← data %>%
left_join(temp, by = c(“YEAR”, “SERIAL”, “PERNUM”)) %>%
mutate(own_kids_u14 = if_else(is.na(own_kids_u14), 0, own_kids_u14)) %>%
filter(own_kids_u14 > 0)
Thanks for sharing this code with other IPUMS users and for noting that there is an issue with our sample code. We will review and modify it as necessary.