Leo Röhlke

What happens when kids get their first smartphone?

They spend less time watching TV/movies, but no less time sleeping or engaging in enrichment activities like reading or going to museums or playing musical instruments or…

From @leoroe.bsky.social

www.sciencedirect.com/science/arti...

#psych

You may have heard of the back-door criterion and the front-door criterion for causal identification.

Introducing: The door-in-the-face criterion (start with outrageous causal claim and hedge) and the foot-in-the-door criterion (start with trivial causal claim and escalate).

Thanks a lot!

Not surprised by the results? Good for you! But the notion of harmful displacement is still alive and kicking, see e.g., a recent OECD report on children, screens and wellbeing. So far, we did not have a lot of robust evidence to show where adolescents take all the time to use their phones instead.

This study is based on amazing longitudinal time-use data from Australia. Using weighted DiD, I analyzed how early adolescents‘ (ages 10-13) time use changes after they receive their first own mobile phone

Happy to share my new study, publ. in Social Science Research 🎉 Bottom line: Yes, adolescents use their phones a lot 🤳 But I find no support for the notion that this use displaces reading, homework, hobbies, physical act., sleep... Instead: Changing media preferences 📺 ->📱. doi.org/10.1016/j.ss...

Being a social scientist means holding two truths at once: 1. This is all deeply complicated 2. I have to submit something by Tuesday

Conni geht zu weit

Social media use in adolescents with and without mental health conditions - Nature Human Behaviour Using a nationally representative UK sample of adolescents with clinical-level mental health symptoms, this Registered Report examined differences in social media use. The results suggest that adolesc...

After 3 years in the making, our Registered Report is out in @nathumbehav.nature.com!

We analyse nationally representative UK data from 3,340 adolescents (aged 11–19) to examine how social media use differs between those with and without mental health conditions. 🧵

www.nature.com/articles/s41...

We Now Know How AI ‘Thinks’—and It’s Barely Thinking at All The vast ‘brains’ of artificial intelligence models can memorize endless lists of rules. That’s useful, but not how humans solve problems.

We Now Know How AI ‘Thinks’—and It’s Barely Thinking at All

Maybe you've heard that AIs are "black boxes"

But a growing body of research keeps arriving at the same conclusion: Today's AIs all work in surprisingly similar -- and simplistic -- ways

1/2

www.wsj.com/tech/ai/how-...

Some cool German/Swiss postdoc opportunities for a quantitative sociologist or similar in this thread:

"Very thrilled to share my new working paper on..."

thanks, Michal!

Screenshot of the title and the abstract of the scientific paper which is referenced in the post. Title: Socioeconomic disparities in Swiss children's use of digital technology: A typological approach based on parental reports". Journal: Journal of Children and Media.

🎉 Happy to share my first PhD paper, published in @journalcam.bsky.social!

💡 4 types of Swiss children’s ICT use 🧒📱
💡 Higher-SES: Limited use vs. learning-oriented use as competing strategies
💡 Lower-SES: Parents more critical of ICT, but limited use is still less common

doi.org/10.1080/1748...

Finally an AI model full of self-doubt, constantly requiring external validation, and taking several months to write a single paragraph.

Sciences Po campus Saint-Thomas, Paris (France)

‪Best wishes! Interesting opportunity to join us in 2025: Sciences Po - CRIS (Paris) is hiring an Assistant Professor (Tenure track)! Candidates with a recent PhD dealing with digital inequalities with a solid methodological background + an ambitious research agenda www.sciencespo.fr/osc/sites/sc...

Are young men and women in Europe becoming more polarized in their political ideologies?
It depends. Across 32 countries:
14 show no meaningful gender gap
7 have a stable small to medium gap with women leaning more left
11 show widening modern gender gaps
osf.io/preprints/os...

Street in Paris, decorated with lights

10 people standing in line for a group picture

Christmas-y vibes in Paris 🎄✨ Was a pleasure to present results from my collab w/ @rsmloh.bsky.social at the Digital Inequalities Symposium at @cris-sciencespo.bsky.social Super cool to meet other early-career researchers interested in this fascinating topic and a great way to end the year!

Ausprobiert und für sehr gut befunden :) Danke!

As a sociologist studying youth and media, I‘d love to be added!

How effective have schools’ ICT resources been—not just in building students’ #digitalskills, but also in narrowing digital skill inequalities? Well... I bring good news, a challenge, and a hopeful twist. 🔍
More in our latest article (open access in Computers & Education!): doi.org/10.1016/j.co...

📄 New working paper!
We explore ICT-related career choices using Swiss longitudinal data.

💡 Key findings:
Girls’ choices align with their ICT interest, boys’ with their ICT self-concept. ICT interest does not predict choice of ICT-specialist careers.

🖥️ Implications for gender disparities in tech!

Would love to be included!

I‘m a sociologist who would like to be added - thanks!

#ECSR2024 was amazing! Great vibe and high-quality research. And, of course, paella, sunshine, and all that rumba 🎶 Looking forward to next year‘s edition! #sociology

Great question! I tested this using a not-too-strict definition of little screen time (n gets too small otherwise) & found no displacement effects (neither screen nor other activities). Maybe cause these kids don't use the phone intensively either? Will consider this aspect for future revisions, thx

Thanks! Indeed, amazing data, but unfortunately pretty unique at the moment

New working paper out! I test whether phone use in early adolescence displaces academically beneficial activities, using DiD and longitudinal #TimeUse data 📱👧👦

Key finding: No time displacement effects on enrichment, sports, & sleep, but reduction in TV/movie watching.
dx.doi.org/10.48350/199...

Womit man es so zu Tagesschau.de schaffen kann, erstaunlich

Und der statistische Test ist ja auch gar nicht ausreichend, die Veränderungen innerhalb der Gruppen wurden ja gar nicht gegeneinander getestet. Was besonders schlimm ist, weil hier ja offensichtlich regression-to-the-mean zu beobachten ist