Guest post from the SIG training reps: Susan Jarvis and Duncan Proctor.
Thanks to all everyone who filled out our survey about quantitative training needs. We had responses covering a wide range of topics suggesting there are all sorts of training requirements out there. We’ll be using this information to plan training courses for next year covering some of the most popular suggestions. Some of the topics suggested are already covered by a number of training courses so over the next few weeks we’ll be putting out a couple of posts with suggestions for existing courses. Today we just wanted to highlight the top three areas based on the number of responses responses. Here are the data if you want to take a look.
Our survey suggested that lots of you are looking for courses in mixed modelling. We were quite surprised about this as we had assumed that this need was mostly covered by universities and research centres but our results indicate there is still a lack of suitable courses. Part of this may be because many mixed modelling courses stop at linear mixed models but, as ecologists, we usually need to use generalised linear mixed models to deal with non-normal data. GLMMs can be much more complicated but there is information out there to help. This GLMM FAQ page is particularly useful with a lot of information in a handy question-and-answer format which is easily to scroll through for what you need. The Highland Statistics team also now have a book covering GLMMs and run courses based on the book material.
Bayesian statistics was also popular, suggesting lots of you haven’t been able to find an introductory course. To get a really good introduction to both Bayesian methods and WinBUGS (some useful software) Marc Kery and Michel Schaub have both written some accessible books for ecologists and also run courses covering the book material.
A few upcoming courses we found include:
- 14th July, University College London, Introduction to Bayesian analysis (not ecology specific).
- 30th-31st July University of Bristol, Bayesian methods to estimate species divergence times.
- 7-9th Sepember, University of Gottingen, Introduction to Bayesian statistics using BUGS and JAGS.
- 26-31st October, Glasgow, Applied Bayesian modelling for ecologists and epidemiologists.
- 3-4th November, University of Reading, Bayesian analysis made easy (not ecology specific).
- 22-26th Feb 2016, University of Strathclyde, Bayesian methods to fit statistical models in environmental science.
Courses on managing and manipulating data were also a popular suggestion. We think this probably reflects both increasing sizes of datasets and the greater recognition for the need to manage data for reproducibility. Many journals will now expect you to archive raw data to supplement publications strengthening the importance of keeping a well-managed dataset that can be picked up and used by other people. There is now a free guide from the BES on data management which is a great place to start but as yet we’ve not found any courses for ecologists. Let us know if you find one!