Anna Langener

Participants of the MITNB workshop, where we stand outside in a group of around 40 people

1st day of #MITNB meeting at @tesc-tilburgu.bsky.social. Excited for the week ahead, where we'll tackle measurement issues in #ESM, e.g. modeling processes across timescales, building a formal theory on measurement, and evaluating statistical assumptions in #ESM data.
Go teamwork <3

‼️job alert ‼️
@emilypresseller.bsky.social is an inspiring researcher and a fantastic mentor-- her lab will be such an incredible launching pad for a budding future scientist!!!

Call. Text. Chat. 988 is here for you - anytime, anywhere. The 988 Lifeline provides judgment-free, compassionate support for anyone experiencing: emotional distress, mental health challenges, problems with substance use, loneliness, thoughts of suicide or self harm. You'll be connected with a caring, skilled 988 counselor who can provide culturally competent support.

Why 988 matters: the 988 Lifeline connects people to 988 counselors across a network of more than 200 local crisis contact centers, offering support rooted in local communities.

The 988 Lifeline isn’t just for thoughts of suicide. 988 Lifeline crisis counselors provide 24/7/365 support for any type of emotional distress. Get compassionate help anytime, anywhere. Call or text 988 or chat online at 988Lifeline.org.
#988Day

AMPPS Call for Papers: Replicability and Reproducibility in Methodological Research. Proposals due September 15. @jkflake.bsky.social 

Are you working on issues of replication and reproducibility of methodological research in psychology? Consider submitting a proposal to my call for papers, happy to chat about your ideas or answer any questions! www.psychologicalscience.org/publications...

Want to learn about dynamic modeling for daily diary, experience sampling, ecological momentary assessment data? 😎

Register for our online course ‘Modeling the dynamics of intensive longitudindal data’ which starts in October 2025! 🤩

utrechtsummerschool.nl/courses/data...

Blog posts | ESM Item Repository

New blog post! 📣 Ever wanted to do more with your open-text experience sampling items, but were unsure 🤔 how to code the responses? @annalangener.bsky.social & @mariestadel.bsky.social show us how it's done with their new Shiny app for open-text ESM data: esmitemrepositoryinfo.com/blog-posts#M...

Image with three main boxes with recommendations for applied researchers on conceptualization, estimation, and interpretation when using network features. Conceptualization: - Discuss why a network feature should be relevant for a given treatment or outcome - Consider plausible effect size and sample size needed to detect an association with a distal outcome Estimation: - Carefully choose preprocessing and network estimation methods - Follow good practices for predictive models (e.g., cross-validation, out-of-sample validation) Interpretation: - Consider uncertainty in node selection and network feature regression - Compare network features with simpler time series features (e.g., person-specific mean or SD)

Can we use features of dynamic networks (e.g. centrality) to improve treatment selection and outcome prediction?

New preprint on the topic: We highlight the role of uncertainty & introduce a Bayesian multilevel approach for uncertainty quantification of network features 🧵
osf.io/preprints/ps...

PhD Gut Feeling or Algorithm? Predicting Patient Outcomes in Critical Care PhD Gut Feeling or Algorithm? Predicting Patient Outcomes in Critical Care (0.8-1.0 FTE) Do you want to cont

Looking for a PhD position? Come work with me! 🤩
In this PhD project, you will study how decisions in medicine (at the intensive care) can be improved using tools like algorithmic advice in both academic and hospital settings.
Talking Dutch is a plus! 😊
www.academictransfer.com/en/jobs/3514...

7/ Want to learn more? Join me for a workshop at the SIPS online conference, where I’ll dive deeper into these topics! ✨ #SIPS2025

6/ To address these pitfalls, we present recommendations for aligning validation and evaluation strategies with the intended use case scenario and created a tool to help researchers investigate whether their strategies and goals are misaligned: annalangener.shinyapps.io/Justintime/

5/ Third, selecting appropriate baseline models is key. Some models may look effective (e.g., AUC = 0.77) but actually underperform compared to simple baselines (e.g., AUC = 0.96). ⚖️

4/ Second, ensuring adequate variability in the outcome variable is crucial. If outcomes are stable, frequent predictions may offer little practical benefit for JITAI.

3/ Centering predictor variables within individuals can improve within-person accuracy but may reduce overall performance.

2/ First, models may perform well overall (AUC = 0.77), but their ability to predict within-person change can be much lower (AUC = 0.56, SD = 0.11). For JITAIs, this will prevent the model from identifying intervention delivery moments and will only discriminate between people.

1/ Many researchers are focused on building prediction models for JITAIs. But a major challenge is the mismatch between model development, evaluation, and application. We use simulations to illustrate three pitfalls.

✨ Excited to share our new preprint: osf.io/preprints/os... ✨

In this paper, Nick Jacobson and I dive into the challenges researchers face when developing prediction models for just-in-time adaptive interventions (JITAI). 🧵👇

#EMA #JITAI #CrossValidation #PassiveSensing

5/ Third, selecting appropriate baseline models is key. Some models may look effective (e.g., AUC = 0.77) but actually underperform compared to simple baselines (e.g., AUC = 0.96). ⚖️

4/ Second, ensuring adequate variability in the outcome variable is crucial. If outcomes are stable, frequent predictions may offer little practical benefit for JITAI.

3/ Centering predictor variables within individuals can improve within-person accuracy but may reduce overall performance.

2/ First, models may perform well overall (AUC = 0.77), but their ability to predict within-person change can be much lower (AUC = 0.56, SD = 0.11). For JITAIs, this will prevent the model from identifying intervention delivery moments and will only discriminate between people.

1/ Many researchers are focused on building prediction models for JITAIs. But a major challenge is the mismatch between model development, evaluation, and application. We use simulations to illustrate three pitfalls.

7/ Want to learn more? Join me for a workshop at the SIPS online conference, where I’ll dive deeper into these topics! ✨ #SIPS2025

6/ To address these pitfalls, we present recommendations for aligning validation and evaluation strategies with the intended use case scenario and created a tool to help researchers investigate whether their strategies and goals are misaligned: annalangener.shinyapps.io/Justintime/%....

5/ Third, selecting appropriate baseline models is key. Some models may look effective (e.g., AUC = 0.77) but actually underperform compared to simple baselines (e.g., AUC = 0.96). ⚖️

4/ Second, ensuring adequate variability in the outcome variable is crucial. If outcomes are stable, frequent predictions may offer little practical benefit for JITAI.

3/ Centering predictor variables within individuals can improve within-person accuracy but may reduce overall performance.

2/ First, models may perform well overall (AUC = 0.77), but their ability to predict within-person change can be much lower (AUC = 0.56, SD = 0.11). For JITAIs, this will prevent the model from identifying intervention delivery moments and will only discriminate between people.

1/ Many researchers are focused on building prediction models for JITAIs. But a major challenge is the mismatch between model development, evaluation, and application. We use simulations to illustrate three pitfalls.

7/ Want to learn more? Join me for a workshop at the SIPS online conference, where I’ll dive deeper into these topics! ✨ #SIPS2025