|
Résumé :
|
We give a short introduction to the principles of Bayesian data analyses and guide the reader through two examples using R. Differences and similarities to frequentist statistics are explained. Bayesian analysis is combining prior information (e. g. from the literature) with information in the data to obtain a posterior distribution of the parameter of interest. The prior and the posterior distributions are probability density functions that describe mathematically what is known of a parameter prior to and after having collected the data. In the first example, we explore the posterior distribution of a proportion of males in a population. In the second example, we estimate the difference in mean wing length between the sexes of a songbird species once without prior information and once using information from the literature as prior information. For both cases, we calculate the posterior probability of the hypothesis that the average wing length in males is at least 1 mm longer than in females. For these examples, we use the R-packages “arm” and “rstanarm”. We explain why the results from a Bayesian data analyses are often easier to interpret and more useful than the result from a classical null hypothesis test. Finally, we recommend some introductory Bayesian text books. The purpose of this article is to help ornithologists understand Bayesian analyses and to provide an easy start-up for doing own Bayesian analyses with R.
|