205067 The Stats Club (S) (WiSe 2010/2011)

This seminar is designed as a regular meeting for discussing and dissecting applied biostatistical problems. It offers an on-demand, goal-oriented, hands-on approach to experimental design and statistical treatment of ecological and evolutionary data from the daily scientific work of the participants. Questions that arise before (preferred) or after experiments are carried out, can be handed in and will be disseminated among participants before meetings take place. This will provide background for a joint discussion of the issue fuelling insight by the collective statistical expertise of all participants.The Stats Club may just as well be used as a forum for reviewers that are unsure whether the statistical treatment, that is presented in a submitted manuscript or a thesis, is appropriate. Questions may include: What exactly is it that I want to know? What kind of data should I collect and exactly how? Given limited sample sizes, will the statistical power be sufficient to allow well-supported conclusions after my study? Is there dependency in my data and how can I control the risk of pseudoreplication? Which is the appropriate statistical model for my type of data? Which error distribution reflects best the underlying stochastic processes? Which variables should be considered as fixed versus random effects terms in my model? Should interaction terms also be considered? How should I interpret the output from my stats software and what conclusions are supported by it?
Discussions in this seminar may be in German or English.

Requirements for participation, required level

Requirements: Participants should have some knowledge of the concept of statistical terms and tests and standard statistical methods (e.g. regression, ANOVA). This seminar is not intended to replace a systematic biostatistics course.

Bibliography

1. Ruxton, G.D. & Colegrave, N. (2006): Experimental Design for the Life Sciences. Oxford University Press, Oxford.
2. Crawley, M.J. (2005): Statistics: An Introduction Using R. John Wiley, New York(see also http://www.imperial.ac.uk/bio/research/crawley/statistics).
3. The R Project for Statistical Computing @ http://www.r-project.org.