Statistic course -new structuring!

9 00 – 10 30 welcome introduction of participants ordinary linear regression

10 30 – 10 45 coffee break

10 30 – 12 00 one-way ANOVA ANCOVA

12 00 – 13 00 lunch break

13 00 – 15 00 how to quantify statistical uncertainty?

15 00 – 15 15 coffee break

15 15 – 17 00 practical work At the end of the day, participants - can apply an LM to own data

17 00 – 18 00 I am here for answering questions concerning your own data

Prerequisites, preparation for the course and handouts

The course is not an R-programming course. You will have more of the course, if you are already are familiar with basics in R. I highly recommend getting familiar with using R before the course, e.g. https://intro2r.com/index.html.

The course is a training of statistical thinking. As a preparation, I recommend reading the lecture notes of the first day: https://tobiasroth.github.io/BDAEcology/basics.html.

The course will follow the book Bayesian Data Analysis in Ecology Using Linear Models with R, BUGS, and Stan by Korner-Nievergelt et al. (2015) Elsevier. I will bring a copy of the book for each participant.

Beside the above book, some material is presented in the draft of the online book:

https://tobiasroth.github.io/BDAEcology/. But most of the course material is still in the printed book.

It is helpful to bring own data to work on. Instructions for preparation below.

Computer and software

You should bring your own computer with pre-installed R (version 4.0 or higher).

R Studio:

It is not mandatory to install R Studio but it is highly recommended. R Studio also recognizes the Stan language and may help with programming Stan. https://www.rstudio.com/products/rstudio/download/

R installation:

To download and install the latest version of R go to www.r-project.org, click on CRAN and then choose a mirror that is close to you. Then, follow the instructions related to your system. It is fine to install R with all the default settings.

To be safe, please, install the following R-packages prior to the workshop:

arm, rstanarm, brms, blmeco, coda, sp, gstat.

Own data project

During the practicals you can either work on your own data or use our example data sets. On Thursday, you analyze your own data. Therefore, bring own data. Please, restrict your own data project to one small question that can be answered using linear models. Thus, you should have a question in the form “How is a response variable y related to a predictor variable x?” and maybe you have a few other (predictor) variables you would like to correct for. Make sure your data is prepared for the analyses and, please, choose a rather small data set (i.e. below 5000 observations (rows), better below 1000). For larger data sets, model fitting takes too much time and impedes you from playing around with different types of models.

Contact me in advance, if you are unsure whether your data is suited for the course. On Thursday afternoon, every participant shortly presents her/his project in a 3-5min presentation.