language spoken at home is such an important topic. new data from @uscensusbureau contains fantastic detail at the national level. otoh, the state level release is underwhelming, much less detail than we got in 2015. https://t.co/A3lULhioUg
since the 2020 census dhc doesn't include a legacy format with geo header file (afaik), here's a stata script that will create a crosswalk for census blocks to cities (useful when you don't want CDPs!)
https://t.co/tGzQjD8Il4
instead, the USCB should use write-ins only to code those who decline to provide a race/ethnicity. The combined "identity" concept is a good one (as Ken Prewitt has argued for decades), but it's subverted by the coding practices in the last census (and ACS).
glad to see USCB called out for their confusion about race, origins, nationality, ethnicity, and ancestry. Great paper that highlights the inaccuracy (and injustice) of coding people into race boxes that they did not select!
glad to see USCB called out for their confusion about race, origins, nationality, ethnicity, and ancestry. Great paper that highlights the inaccuracy (and injustice) of coding people into race boxes that they did not select!
USCB has added many new items to their race/ethnicity list (used to crosswalk free text entry to numeric codes)! And also now includes "multicodes" like Dougla & will code to Black *&* Asian from that. Should Mestizo also code to specific multiple races? https://t.co/7YTWtr3udm
I will give a one-day short course on Subnational Bayesian Population Projections on April 10 at the 2025 @PopAssocAmerica annual meeting https://t.co/80JCWAEXu3 w Hana Ševčíková: https://t.co/QEpuqLjCRc
some folks have asked about how to generate CI for 2020 Census data -- ESRI published a tool for that, and now there's an official source. https://t.co/GhqwDeUZxC
Pleased to report that our last school enrollment forecast error for the fall 2024 kindergarten class overshot by just 154 students (+0.35%) -- that sets the expectations for our next forecast quite high!
Initial enrollment data from Portland Public Schools for the new school year shows the number of students dropped for a third consecutive year, and the numbers are slightly lower than expected. https://t.co/oOyBLzwni5
The Graduate Group in Demography at the University of Pennsylvania is now accepting applications for its PhD program in Demography. We look forward to recruiting a strong and diverse cohort for Fall 2025! Applications are due December 15, 2024. Learn More: https://t.co/O7uM3xMN53
after a census, political districts should be redrawn without dividing 'communities of interest', but how to define those? New paper by PRC Research Associate David Swanson contributes some ideas on how to make the process more transparent. https://t.co/kiV8L8uLjF
new paper finds that census's top-down algorithm (TDA) was crucial in making differential privacy work for the 2020 US Census. Parameters used in 2020 Census provided accuracy in most cases similar to swapping https://t.co/hYIUYtqwC0
updates to the federal race/ethnicity standards are exactly as expected-- nothing groundbreaking. A combined race/ethnicity question + MENA added, plus minor tweaks to language. A step forward, but I don't think it gets us where we need to be for 2030. https://t.co/MV5QVhDMJ7
D. Hofstadter: "the toughest challenge facing AI workers is to answer the question: What are the letters 'A' and 'I'?" (1995). Nice that cursive reinforces letters as ideas and not fixed shapes. We see this when reading (fonts), less in writing/typing. (https://t.co/0qZq5EEXLZ)