Intro to Bayesian Modeling
Introduction
LIvestream or asynchronous course by instats, Jan 2025
INTEREST CATEGORY: MARKETING RESEARCH
POSTING TYPE: Events
Posted by: ELMAR Moderator
This 5-day workshop provides an intuitive hands-on introduction to statistical modelling, viewed from the Bayesian perspective. The course starts by covering the very basics of what a model is, building up to fairly sophisticated models by the last session (for example, predicting COVID-19 outcomes from biomarker data). An official Instats certificate of completion is provided at the conclusion of the seminar. For European PhD students, the seminar offers ECTS Equivalent points.
The workshop is led by Dr. Rosina Savisaar, an evolutionary geneticist affiliated with Mondego Science, offers a gentle introduction to the world of Bayesian statistical modelling using R. It is open both to participants who are new to regression modelling in general, as well as to those that already have experience with classical (frequentist) statistical modelling and want to re-imagine statistics and scientific inferenence from a Bayesian perspective.
In recent years, it seems like the Bayesian way of doing statistics is everywhere. This is due in large part to the powerful computational methods that come with it. These methods allow researchers to fit very complex models that can sometimes be problematic for more classical approaches. Another benefit of the Bayesian framework is that it provides an alternative to P-values: a concept that is becoming increasingly controversial, with many (e.g. the American Statistical Association) claiming that its use promotes poor scientific practices. Finally, one may also argue that Bayesian methods more closely mimic our natural intuitions about science than do frequentist statistics. In the Bayesian world, a researcher can formulate a hypothesis and directly calculate the probability that this hypothesis is true – a very direct answer to the scientific question asked. In contrast, in the frequentist paradigm, hypotheses are typically tested using the P-value – a very indirect metric of the validity of the hypothesis.
Full information available here: