Note also the blog post by Daniel Lakens on the same topic here. Update March 23, 2022: a more concise / slightly modified version of these thoughts was published as a letter in TREE, co-authored together with Fred Barraquand. Background Criticism of p-values has been rampant in recent years, as were predictions of their imminent…
In regression analysis, a common problem is to decide on the right functional form of the fitted model. On the one hand, we would like to make the model as flexible as possible so that it can adjust itself bias-free to the true data-generating process. On the other hand, the more freedom we give the…
The central quantity in Bayesian inference, the posterior, can usually not be calculated analytically, but needs to be estimated by numerical integration, which is typically done with a Monte-Carlo algorithm. The three main algorithm classes for doing so are Rejection sampling Markov-Chain Monte Carlo (MCMC) sampling Sequential Monte Carlo (SMC) sampling I have previously given…
As announced a while ago, we had moved our now already traditional summer school in Bayesian Statistics to Bergen, Norway this year. Maybe fitting for such a course, the weather turned out to be very different from the long-term frequency, in what must be the upper 1% quantile of sun intensity for the region at…
I’ve just returned from two weeks in France, the first week on the International Statistical Ecology Conference 2014 in Montpellier, and the second at the Laboratoire d’Écologie Alpine (LECA) in Grenoble, visiting the groups of Wilfried Thuiller and Sébastien Lavergne, which was both great. Some impressions from the ISEC: First of all, my compliments to…
Approximate Bayesian Computation (ABC) is an umbrella term for a class of algorithms and ideas that allow performing an approximate estimation of the likelihood / posterior for stochastic simulation models when the likelihood cannot be explicitly calculated (intractable likelihood). To give you the idea in a nutshell: to approximate the likelihood, consider that for a…
Since a few years, various subsamples of Jörn Pagel, Joseph Chipperfield, Björn Reineking and I have been teaching an introductory summer school in Bayesian statistics. In previous years, we were in Bayreuth 2011 Bayreuth 2012, Göttingen 2013 and Freiburg 2013. This year, we will be in a beautiful field station close to Bergen, Norway. The…
We are looking for candidates to fill a 3-yr postdoc position in our lab. The plan is to use Bayesian statistics to connect a process-based forest model to data, and use the fitted models to test the effect of functional diversity on ecosystem responses and community stability. Details here or via NatureJobs.
The end of last year has been busy, with the result that my blogging activity had nearly ceased. No R holiday cards or end-of-the-year posts like other people. I’m happy to say though that, at least in spirit, but hopefully soon in writing as well, I’m back on track, so more frequent posting will commence…
I recently posted about a reply to a PNAS paper in which we show that, contrary to the claim in the original publication, the “true” model does by no means delivers worse predictions for chaotic population dynamics than time-series methods without any specific mechanism underlying, provided appropriate statistical estimation methods are used. To be clear,…
Biogeosciences just published a discussion paper on a project I have been working on for quite a while now, namely on likelihood approximations for stochastic forest models (see also our review on the topic of simulation-based inference in general). From the technical side, it may be interesting that we don’t use Approximate Bayesian Computation (ABC),…
I just returned from our Summer School on Bayesian Statistics that we held in the mountains around Freiburg. Thanks to all the participants as well as my fellow lecturers Jörn Pagel and Joe Chipperfield for the really amazing time up there in the hut, hope to see you all soon!
… but does “model-free” forecasting really outperform the “true” model? I’ve been busy in the last while, which is both the reason why it has been quiet on this blog and why I didn’t manage to write earlier about a note that we submitted three weeks ago to the arXiv. The note responds to a…
As in the last years, we organize a course in Bayesian statistics this summer. See invitation below, feel free to distribute widely. 3rd International Summer School An Introduction to Bayesian Modeling for Ecologists in Freiburg (Germany) 28th July to 2nd August 2013 Motivation Bayesian inference is an increasingly used statistical framework in ecology, but…
I just returned from teaching a workshop on Bayesian statistics together with Joe Chipperfield in Göttingen. The workshop was organized by the Biodiversity, Macroecology and Conservation Biogeography group in the aftermath of the Macroecology meeting, so quite a few people stayed over the weekend after the meeting, and it generally turned out to be a…
A comment by Deborah Mayo on a recent post by Andre Gelman started a heated and surprisingly subtle discussion on the correct (frequentist) definition of the p-value that currently continues on Larry Wasserman blog. Read for yourself – I for myself am still happy to write p(D>d|H0) ;).
This week’s edition of Nature featured a comment by Drew Purves and colleagues titled “Ecosystems: Time to model all life on Earth”, in which they propose that we need a new model type, general ecosystem models (GEM), that describe the interplay of all major organism types on earth. They say that GEMs […] could capture…
Our new perspective paper on “Connecting dynamic vegetation models to data – an inverse perspective” is now available online at the Journal of Biogeography (see also this post by my coauthor James Dyke). The paper is a part of a special issue on distribution models “The ecological niche as a window to biodiversity” that will…
A few days ago, I came across a very interesting post on “Mechanistic models: what is the value of understanding?” by Amy Hurford on her Just Simple Enough: The Art of Mathematical Modelling blog. The post discusses the old, but for me still somewhat unresolved question of the relationship / definition / statistical treatment /…
Our new review on statistical inference for stochastic simulation models is now available at Ecology Letters. The paper discusses how approximate likelihood functions can be generated from stochastic simulations. Such simulation-based likelihood approximations are a very general and powerful tool because they allow us to apply statistical methods to a large class of models such…
theoretical ecology 