Perline Demange
perlinedemange.bsky.social
Perline Demange
@perlinedemange.bsky.social
Postdoctoral fellow at PROMENTA, University of Oslo.
Genetics – Intergenerational Transmission of Inequalities – Mental Health – Education
https://perlinedemange.github.io/
To end this thread on some concrete ideas to improve the ethical and social implications of social science genetics, here is a compilations of slides from past-month talks by @rosacheesman.bsky.social @evangiangrande.bsky.social and Bennett McIntosh.
June 25, 2025 at 3:10 PM
Implications are of course highly relevant for criminology.
(Yes Trump said this quote, not Shakespeare, as I unfortunately said loudly when Mareile asked... 🥲 Fooled by my simple heuristic: quote in English that looks like a meter = Shakespeare 🥲)
June 25, 2025 at 3:10 PM
"Socially-responsible design starts with the research question" What type of research question we ask, how we ask it, and how we try to answer it is not neutral, whether we like it or not.
June 25, 2025 at 3:10 PM
Rosa went further with the example of "low-PGI children": such language implies genetic essentialism, this is labeling and reductive. To read: www.nature.com/articles/s41...
June 25, 2025 at 3:10 PM
Yaël Friedman (philosopher of science) took the example of geographical-based labels and how they downplay complexity and smuggle back racial assumptions.
Cant remember where, but someone suggested to use "European-like genetic ancestry" instead of "European ancestry".
June 25, 2025 at 3:10 PM
"Genetics isn't just happening in our bodies. It's happening inside **systems**" @ziadaayorech.bsky.social told us about common misconceptions, including our tendency to forget that everything we study is context-dependent.
June 25, 2025 at 3:10 PM
Brilliant panel on the education of mental health professional and the role of intersectionality, as part of Oslo Pride, with PROMENTA's Sam Fluit, Jacob Evje and @ziadaayorech.bsky.social, chaired by @nastjaontheroad.bsky.social.
My highlights 👇
June 24, 2025 at 8:42 PM
Ending the conference with a panel about Mendelian Randomization with @mendelrandom.bsky.social Alice Carter and Dalton Conley. My main take-away: use negative controls!
@essgn.bsky.social
May 23, 2025 at 3:46 PM
@sjoerdalten.bsky.social combines a trio PGI design with the sibling design (by Trejo, the model where you only need the phenotypes of both siblings) to look at the association of personality PGI and SES outcomes in Lifelines and the dutch registry. @essgn.bsky.social
May 23, 2025 at 11:26 AM
Independent contributions to externalizing and internalizing symptoms from common and rare variants shown by @olivia-wootton.bsky.social in MCS and Alspac.
Also always nice to see that the NonCog PGI association is not mediated by cognitive skills 🤞
@essgn.bsky.social
May 23, 2025 at 10:34 AM
Very thorough investigation of the impact of low dose ionization due to nuclear incident in the UK. By Rossella da Sabbato @essgn.bsky.social I’m learning a lot here!
May 23, 2025 at 9:00 AM
Using having two first children of same sex as an IV, Dalton Conley disproves the Trivers-Willard hypothesis: it appears having a girl is taking a bigger toll on parents’ epigenetic clocks. @essgn.bsky.social
May 23, 2025 at 8:38 AM
“Social science genetics will continue as a field but your publications will stay” @libertyruth.bsky.social told us her wishes and advice as an NHB editor.
-> Talk about your limitations, use correlation terminology when not testing causation, document your ethical considerations, write FAQs
May 23, 2025 at 8:29 AM
Interesting method challenge from Niels Rietveld: how do you do an extended twin family model when you don’t know zygocity and have extremely gendered phenotype? They compare same-sex twins and same-sex closely spaced siblings! They run this on entrepreneurship in CBS. @essgn.bsky.social
May 22, 2025 at 3:54 PM
Financial stress is most strongly associated with epigenetic aging amongst a broad range of negative life events (between-family estimates). By Bodine Gonggrijp @ntrbiopsy.bsky.social @essgn.bsky.social
May 22, 2025 at 3:13 PM
In the UKB, breastfeeding (self-report) is associated with adult height and cognitive performance, even within-siblings. By Emil Sorensen. @essgn.bsky.social
May 22, 2025 at 2:58 PM
A different kind of talk by Bennett McIntosh, with an ethnographic point of view. Teaching us about boundary objects and how social science genetics fits into reproductivism. Brilliant! @essgn.bsky.social
May 22, 2025 at 2:55 PM
What did you guys do to me… 😂
May 22, 2025 at 2:44 PM
And that’s already a wrap for me, thank you all for coming to my poster! @essgn.bsky.social #ESSGN2025
May 22, 2025 at 1:05 PM
Now trying to wrap my head around measuring ancestry-based assortative mating. Interesting methods being used by @luyinz.bsky.social. @essgn.bsky.social
May 22, 2025 at 11:27 AM
Associations between psychopathologies PGIs and cognition are the same when controlling for P factor. Within- and between-family association pattern is different depending on the psychopathology and associations with verbal and nonverbal IQ as well. @essgn.bsky.social by Wangjingyi Liao
May 22, 2025 at 10:57 AM
Ok I was a bit quick to post, as deciles plots (bar plots) appear better at conveying info about averages (while scatter plots better for prediction about individuals). Showing both types of plots seems to have basically same effects as showing only decile plots.
May 22, 2025 at 9:06 AM
Kicking off #ESSGN2025 @essgn.bsky.social with @michellemeyer.bsky.social advocating for scatter plots vs bar plots in social science genetics. Everyone (scientists and public) is better at interpreting effect sizes with scatter plots.
May 22, 2025 at 8:54 AM
Using self-reported mental health data and academic standardized tests, we compared siblings and their children to answer: Is the sibling with worse mental health more likely to have a child with lower achievement?
April 16, 2025 at 10:35 AM
New preprint!

We find no evidence that parental mental health influences children's academic achievement when comparing families in the Norwegian MoBa study.

osf.io/preprints/ps...

Quick thread 👇
April 16, 2025 at 10:35 AM