iiinova — where your digital self keeps evolving

Felix Dietrich @felix-dietrich.de · 28/09/2026

Das wäre natürlich das bessere Zitat für den Titel gewesen. Mist!

Felix Dietrich @felix-dietrich.de · 28/09/2026

🔹 Algorithmic personalization is a mixed bag. More algorithmically curated sessions elicited fewer memories and less meaningful experiences, but serendipitous discoveries were enjoyed and appreciated more.

110

Felix Dietrich @felix-dietrich.de · 28/09/2026

🔹 Context matters. Memories were more likely with older songs, songs from listeners’ younger years, their favorite genres, and their all-time top tracks, and less likely when listening with others. Music on the side (e.g., while working) felt less meaningful.

100

Felix Dietrich @felix-dietrich.de · 28/09/2026

🔹 Meaning comes in two flavors. Sometimes it involves effortful reflection on mixed experiences. More often, it’s intuitive: a positive memory simply “makes sense” and is both enjoyed and appreciated.

100

Felix Dietrich @felix-dietrich.de · 28/09/2026

What did we find? 🔹 Biographic resonance is common. Music listening surfaced personal memories in about a quarter of all listening sessions. When it did, listening felt much more meaningful.

100

Felix Dietrich @felix-dietrich.de · 28/09/2026

🎶 New paper in Communication Research! Ever listened to a song and suddenly you’re back at that summer, that road trip, that concert with friends? We examined biographic resonance, when music connects with our memories. 📄 The paper: doi.org/10.1177/0093... 💛 Open Access: osf.io/preprints/ps...

2106

Felix Dietrich @felix-dietrich.de · 06/09/2026

The first tutorial is out now. A related intro paper in Publizistik will be published soon. Today, @dpossler.bsky.social and I presented the tutorial at a method workshop at @dgps.bsky.social in Luxembourg (unfortunately not on a tourist boat, even though the pictures may give this impression).

010

Felix Dietrich @felix-dietrich.de · 06/09/2026

The best thing? You don‘t need deep coding skills to run them—copying and pasting a few lines of code will do it. To help scholars adopt this exciting infrastructure and use it responsibly, we are working on a website featuring beginner-level step-by-step tutorials: llm-content-analysis.com

llm-content-analysis.com

Open Source LLMs for Content Analysis

141

Felix Dietrich @felix-dietrich.de · 06/09/2026

Ready-to-use, open-source Large Language Models (LLMs) are available online (e.g., on Hugging Face). They are free to use, adaptable, scalable, run locally, and achieve a level of performance that can compete with manual content analysis.

110

Felix Dietrich @felix-dietrich.de · 06/09/2026

Asking ChatGPT or Claude about your research is easy — but what about using reproducible and privacy-compliant large language models (LLMs) for analyzing texts? Well, that‘s pretty straightforward too. A thread 🧵⬇️

111

Felix Dietrich @felix-dietrich.de · 18/03/2026

Dev version available on GitHub 👇 pak::pak("felixdidi/mlstats") github.com/felixdidi/ml... Feedback and contributions very welcome! ⭐ the repo if it’s useful to you 🙏 #rstats #multilevel #statistics #openscience

github.com

GitHub - felixdidi/mlstats: Tools for multilevel analysis

Tools for multilevel analysis. Contribute to felixdidi/mlstats development by creating an account on GitHub.

020

Felix Dietrich @felix-dietrich.de · 18/03/2026

Output can be exported as a gt or tinytable object — which renders straight into a Word doc via Quarto. Also included: variable decomposition for REWB models and Bayesian estimation via brms if you want credible intervals instead of p-values.

220

Felix Dietrich @felix-dietrich.de · 18/03/2026

If you are working with multilevel data, try out my new package mlstats 📦 — an #rstats package for multilevel analyses. You can create publication-ready descriptive tables like the one below (including within- AND between-group correlations + ICCs). github.com/felixdidi/ml... Details in thread 🧵↓

Example screenshot of multilevel descriptive table created with mlstats R package

15918

Reposted by Felix Dietrich

hackingcommsci @hackingcommsci.bsky.social · 05/02/2026

✨We’re live! Our updated Hackathon website is online — explore it here for more information: hackingcommsci.org 🎉 You’ll find everything you need to know for the event at the SU School for Data Science and Computational Thinking in Stellenbosch on 3rd–4th of June 2026 🇿🇦

hackingcommsci.org

Home | Opening Communication Science

ICA 2026 Pre-conference Hackathon

032

Reposted by Felix Dietrich

Hadley Wickham @hadley.nz · 06/11/2025

Do you teach #rstats? Do your students complain about how lame and old-fashioned dplyr is? Don't worry: I have the solution for you: github.com/hadley/genzp.... genzplyr is dplyr, but bussin fr fr no cap.

github.com

GitHub - hadley/genzplyr: dplyr but make it bussin fr fr no cap

dplyr but make it bussin fr fr no cap. Contribute to hadley/genzplyr development by creating an account on GitHub.

42460166

Felix Dietrich @felix-dietrich.de · 26/10/2025

Also, the Web API provides a lot of meta data about tracks (more than any other entertainment platform that I am aware of), so this should be useful for both linkage designs like ours, but also when using DDPs.

110

Felix Dietrich @felix-dietrich.de · 26/10/2025

Available papers from this project / using this method (so far... more to come, hopefully soon): dl.acm.org/doi/10.1145/... osf.io/preprints/ps...

110

Felix Dietrich @felix-dietrich.de · 26/10/2025

We used the API for "real time" donations / linkage of recently listened tracks in each ESM questionnaire, no DDPs (see developer.spotify.com/documentatio...). So no experience about average download time, but API worked pretty great (unless you are interested in podcasts, which are not supported).

120

Felix Dietrich @felix-dietrich.de · 15/10/2025

Thank you @overbye.bsky.social — going to Greece to present the paper wasn’t so bad either 👀 Also thanks to my wonderful co-authors @aliciaernst.bsky.social, @rkreling.bsky.social, @aliciagilbert.bsky.social, @mscharkow.bsky.social, @leonardreinecke.bsky.social and Benedikt! More to come soon!

The research group at the coast of Hermoupolis, Syros

182

Reposted by Felix Dietrich

Rebekka Kreling @rkreling.bsky.social · 09/10/2025

Interested in… a) taking theories out of the lab (and try to specify them to work irl) b) computational methods (data donations, LLM content analysis) c) media coping d) all of the above? Then our new HCR-paper is *just* for you! doi.org/10.1093/hcr/...

193

Reposted by Felix Dietrich

Alicia Gilbert @aliciagilbert.bsky.social · 24/09/2025

💡Pub alert 💡 Why do users stop a media activity? Today I presented "101 reasons to stop: A systematic review of media cues for disengagement" at #CHIGreece2025 🏖️ The full paper - the 3rd of my PhD thesis - is now available #openaccess at dl.acm.org/doi/10.1145/... Summary below ⬇️

1266

Reposted by Felix Dietrich

hackingcommsci @hackingcommsci.bsky.social · 13/06/2025

From text-as-data to sleep data collection, this year‘s #hackica25 in Boulder, CO, was a blast and we‘re proud to say that we‘ve (again) successfully hacked communication science. Thanks to all participants and see y‘all at #ica25 ✨

14118

Reposted by Felix Dietrich

Institut für Publizistik (IfP) @ifpmainz.bsky.social · 29/01/2025

#RezFo25 AWARDS 📢🏆 Auf der diesjährigen Tagung der Fachgruppe Rezeptions- und Wirkungsforschung (@rezfo.bsky.social ) der DGPuK wurden alle drei Best Paper Awards von Wissenschaftler:innen des IfP (mit)verfasst. Details 👇

1135

Felix Dietrich @felix-dietrich.de · 20/01/2025

Ever wondered what researchers beyond #commsky have to say about autonomy in digital technology use (e.g., affordances, situations, context)? Check out our interdisciplinary, large scale (computational) scoping review w/ @anishaarenz.bsky.social @leonardreinecke.bsky.social doi.org/10.1093/iwc/...

095

Felix Dietrich @felix-dietrich.de · 17/01/2025

Very fun project with @rkreling.bsky.social, @aliciagilbert.bsky.social, @leonardreinecke.bsky.social. Most interesting challenge for me: How can we utilize the scarce information in Netflix DDPs to study entertainment phenomena in the wild using data donations.

050

Reposted by Felix Dietrich

Alicia Gilbert @aliciagilbert.bsky.social · 01/07/2024

📢1st dissertation paper alert📢 Out now in JoC: Why do people #disengage from media use? Very excited to see this one published🌟 Kudos to my wonderful co-authors @leonardreinecke.bsky.social @adrianmeier.bsky.social Susanne Baumgartner & @felixdidi.bsky.social academic.oup.com/joc/advance-...

093

Felix Dietrich @felix-dietrich.de · 16/03/2024

It was fun to talk about our experiences with using Transformers in #MediaPsych and #Communication research at #dgpuk2024. Great projects with @dpossler.bsky.social, Jule Scheper, Anica Lammers, Rebekka Kreling, @aliciagilbert.bsky.social, @leonardreinecke.bsky.social.

People looking at the poster at the conference

073