Here, I helped look @ startling statement by Democratic Alliance's Alan Winde: "This province is responsible for 78.9% of all jobs in the country in the last 5 years." It's extremely misleading. Eg the following is also true: "KZN accounts for 107% of net job growth in the last 5 years."

At #AAPOR24 tomorrow, Sihle et al investigate differences between self-administered and face-to-face modes in population inferences related to regression coefficients that capture subgroup differences. The study focuses on sensitive health and reproductive measures.

At 9:45 Sihle will present Mode Effects on Regression Coefficients Capturing Subgroup Differences: Comparing Regression Models across the Web-Based American Family Health Study and the Face-to-Face National Survey of Family Growth myumi.ch/RmVE4 #AAPOR2024

Jonathan Mendelson & Michael Elliott provide mathematical theory on how to allocate single-stage designs in a manner that incorporates the effects of nonresponse on cost efficiency myumi.ch/23xNn

👀Americans who experience downward mobility are more likely to attribute inequality to structural processes, while those who rise are more likely to reject structural explanations. NEW: UM Population Studies Center's Sofia Hiltner and Erin A. Cech (she/her) in Social Forces. Full abstract below ⬇️

Yes, we have to find a path forward but such poor reasoned tweets are not helping: Prof Burger et al, have a paper that shows the evolution of racial wage gap in SA, this gap widened post 1994 (discrimination), despite AA policies up to 2003 (BEE introduced).

Only around 50% of academic papers focus on low- and middle-income countries, despite these countries comprising over 80% of the global population. See how World Bank Policy Research Papers help bridge this gap: wrld.bg/uUj950SBzRT

Today's editorial from Business Day. SA needs a census it can count on businesslive.co.za/bd/opinion/202… via BusinessLIVE

💔

Here is a recording of my webinar last week on practical approaches to estimating selection bias in non-probability samples or low RR probability samples. These measures are easy to use and can provide insight re: potential selection bias! Rebecca Andridge youtu.be/WKovy4Nr4tc

At a school he enters wielding guns and pangas and he is still walking as a free man! There is no leadership really 😡

Despite our earlier concerns (published by the SAMRC) that the data from the 2022 SA Census is not fit for purpose, and StatsSA's vocal defence of it, StatsSA's latest publication says the census cannot be relied upon to provide useful data on a number of crucial indicators...

Breaking news 😆: A new study finds members of Congress whose families once owned slaves are substantially wealthier than those whose families did not. #slaveryarchive thehill.com/newsletters/su…