All details in

@RosanaFerrero @ReginaNuzzo I agree although the FDR argument just needs power and a-priori probability of an effect, the other issues come on top of that! In other words, you can easily get 50% FDR even if you work by the book!

Our study "Analyzing longitudinal trait trajectories using GWAS identifies genetic variants for kidney function decline" led by Simon Wiegrebe and Iris Heid just out in @NatureComms

On my way back from a short visit at Uni Birmingham! Thanks @cabral_sarmento for inviting me!

Given the general exodus around me, I want to say that while I will stay here for the moment, you can also find me at #Bluesky under @florianhartig

sensitivity analysis, uncertainty analysis and model inversion for complex (ecological) models. Comments welcome!

@bynyamyn @ReginaNuzzo I‘ll send them to you!

Had a wonderful time this week at the @SFE2_2024 conference in #Lyon. Thanks to Stéphane Dray for inviting me to the symposium "Advances in Statistical Ecology", and great catching up with @efronhofer @guiblanchet3 @FredBarraquand and many more #SFE2

@strengejacke No I wouldn‘t have expected any changes in the residual values … could you provide more context, ideally via our issue tracker at GitHub?

To find out more about DHARMa, visit

RT @aubreyclayton: Doing some Bayesian statistics and confused about which test to apply in which situation? Just consult this handy flowch…

Fantastic chance to present much of our recent work on the connection of ML and #DeepLearning with statistics, including the cito #Rstats package for building neural works with R as well as our recent work on inductive biases in different ML algorithms