Round-up: Meta-analysis of the black–white IQ gap

The most interesting articles that came to our attention this week

Investigating the replicability of the social and behavioural sciences. Andrew Tyner and colleagues attempted to replicate 274 separate findings from papers published in social science journals from 2009–2018. Overall, 55% of the claims replicated in the sense of obtaining a statistically significant result. The average Pearson correlation was r = .25 in the original studies and r = .10 in the replication studies. (Note that there is a companion paper on reproducibility.)

Tweet by Ryan Burge. The author examines ideological differences between mainline clergy and their congregants in the US. Analysing data from the Cooperative Election Study, he finds that clergy are far more liberal than their congregants. Among clergy, 45% identify as liberal Democrats versus 12% who identify as conservative Republicans. Among laity, 24% identify as liberal Democrats versus 37% who identify as conservative Republicans.