Per Engzell
@pengzell.bsky.social
Interested in how the rich stay rich and the poor poor. Sociologist at @sriucl.bsky.social @ucl.ac.uk. He/him/his. http://perengzell.com Photo bomber @simoneschneider.bsky.social
50. Prince: ieeexplore.ieee.org/document/667...
November 30, 2025 at 5:17 PM
50. Prince: ieeexplore.ieee.org/document/667...
49. Led Zeppelin: link.springer.com/article/10.1...
November 30, 2025 at 5:13 PM
49. Led Zeppelin: link.springer.com/article/10.1...
48. The Durutti Column: www.themoonlight.io/en/review/in...
November 30, 2025 at 5:11 PM
48. The Durutti Column: www.themoonlight.io/en/review/in...
47. Grace Jones: www.oreilly.com/library/view...
November 30, 2025 at 5:03 PM
47. Grace Jones: www.oreilly.com/library/view...
46. Cornelius: link.springer.com/article/10.1...
November 30, 2025 at 4:58 PM
46. Cornelius: link.springer.com/article/10.1...
This sent me down a rabbit hole on domestic violence legislation which of course was shockingly late and often passed for the wrong reasons (nuisance to neighbors)
November 30, 2025 at 12:35 PM
This sent me down a rabbit hole on domestic violence legislation which of course was shockingly late and often passed for the wrong reasons (nuisance to neighbors)
Good story but not a true story en.wikipedia.org/wiki/Rule_of...
November 30, 2025 at 5:24 AM
Good story but not a true story en.wikipedia.org/wiki/Rule_of...
Biggest culture shock in moving to Britain was learning that that passes for an argument
November 30, 2025 at 5:12 AM
Biggest culture shock in moving to Britain was learning that that passes for an argument
Reposted by Per Engzell
Really baffling that two most common uses of multiverse analysis are: 1. Run a bunch of correlated analyses & use correlation as evidence for the strength of inference, or 2. Run a bunch of different analyses under wildly different model (mis)specifications to show that the 'effect' is not reliable.
November 29, 2025 at 4:46 PM
Really baffling that two most common uses of multiverse analysis are: 1. Run a bunch of correlated analyses & use correlation as evidence for the strength of inference, or 2. Run a bunch of different analyses under wildly different model (mis)specifications to show that the 'effect' is not reliable.