I want to examine racial/ethnic changes in public vs. private school enrollment for grades 1-12 in Chicago, 1990 to 2013. But it appears that GRADEATT is not available for 1990 (https://usa.ipums.org/usa-action/vari…), even though this variable is available for every other year since 1960, and even though 3 basic categories of grade level (preprimary, elementary + high school, and college) are available via the 1990 FTP download site (which doesn’t let me cross-tabulate with school type & race). If there is indeed no GRADEATT data for 1990, I would just like to know why not, if someone could please explain. Thank you!
In the 1990 Census, respondents were asked to report the highest grade level completed (Question 12), which is recorded in the variable EDUC. Unlike with the other samples from 1960 to present, the 1990 Census did not ask respondents if they were currently attending (or had recently attended) school. As a result, it is not possible to provide the variable GRADEATT for the 1990 sample.
Hope this helps.
Thanks. Then, I wonder how a different tabulation of this information can be made available via the FTP download site (http://www2.census.gov/census_1990/CD…). The list of tables shows a breakdown of preprimary, elementary and high school, and college enrollment. Do you think they used educational attainment, school enrollment status (yes/no) and age do construct an estimate?
According to Appendix B of the 1990 Census Report on Education, school enrollment was determined using the answer to Question 11 (SCHOOL) and level of school enrollment was derived using Question 12 (EDUC). User Note 4 of the Census Report on Education provides a crosswalk from Question 12 (EDUC) to the derived enrollment levels found in your Census tables.
Hope this helps.