ISBE2016 Post-conference Symposium
Statistical quantification of individual differences: educational and statistical approaches informing study design, analysis and inferences of multi-level behavioural data
Niels J. Dingemanse1,2* (email@example.com)
Yimen G. Araya-Ajoy1 (firstname.lastname@example.org)
Denis Réale3 (email@example.com)
Participation is open to all scientists registered to the ISBE2016 conference. Please drop Yimen Araya-Ajoy (firstname.lastname@example.org) an email if you are considering to attend the symposium; this will give us an idea of how many people to expect.
1Behavioural Ecology, Department of Biology, Ludwig-Maximilians University of Munich (LMU), Planegg-Martinsried, Germany
2Research Group Evolutionary Ecology of Variation, Max Planck Institute for Ornithology, Seewiesen, Germany
3Center for Biodiversity Dynamics, Department of Biology, Norwegian University of Science and Technology (NTNU), N-7491 Trondheim, Norway
4Département des Sciences Biologiques, Université du Québec à Montréal, Canada
Variation exists at all levels of biological organization: among species, within-populations among-individuals, and for labile traits, within-individuals. Mixed-effects models represent ideal tools to quantify multi-level variation and are increasingly used, for example in studies of animal personality and behavioural plasticity. Mixed-effects models are complex and two issues hamper their proper usage:
i) the relatively few educational resources to gently teach students how to implement and interpret them, and
ii) the lack of tools ensuring that statistical parameters of interest are correctly estimated.
We focus on two key issues. First, we will highlight educational tools that researchers (teachers and students) may apply to gently learn how to use the mixed-effects model for addressing a wide range of questions. Second, we will discuss the importance of the study’s sampling design in estimating parameters of biological relevance accurately, precisely, and with sufficient statistical power. For example, what is the optimal sampling design for a given question given constraints imposed by the study system? Such questions are faced by every researcher and determine the reliability and interpretation of the data, as well as reviewer support in the publication process. The symposium will thereby highlight the importance of simulation-based tools for education and research purposes.
Niels Dingemanse (Ludwig Maximilian’s University of Munich)Introducing the symposium and the mixed-effects modelling framework
Hassen Allegue (University of British Columbia)
Introducing the educational and statistical tool “SQuID”
Denis Réale (Université du Québec à Montréal)
Biased sampling schemes causing pseudo-repeatability in behaviour
Raphael Royauté (North Dakota State University)
Sampling requirements for testing differences in repeatability between datasets
María Moirón (Max Planck Institute for Ornithology)
Optimal sampling designs for studying of state-behaviour feedback loops
Dave Westneat (University of Kentucky)
Counter-intuitive effects of mean-centring in mixed-effects model analyses
Samantha Patrick (University of Liverpool)
Quantifying autocorrelation across temporal scales to reveal hidden trade-offs
Barbara Class (University of Turku)
How to avoid bias in estimates of assortative mating in wild populations?
Kate Laskowski (Leibniz Institute of Freshwater Ecology & Inland Fisheries)
Year-round telemetry reveals extensive individual variation in wild fish
Yimen Araya-Ajoy (Norwegian University of Science and Technology)
Assessing support for non-zero variance components in Bayesian mixed models
Shinichi Nakagawa (University of New South Wales)
Closing discussion, future perspectives, and wrap-up
Currently no messages in this area.