Do you think the complexity of biological data is beautiful? Does variance partitioning get you excited? Do you toss and turn at night thinking about ways to optimize that one little piece of code? Do you like sharing this excitement with other folks and figuring out tools to do so?
Well then I have the opportunity for you! Come join our SQuID working group as a postdoctoral scholar. Not only is this a super fun, highly collaborative working group on the cutting-edge of statistical tool development, but also we have a really great name. While the postdoc will officially be stationed out of NTNU in Norway there will be lots of flexibility for remote possibilities. See the ad below (official application will open in a few weeks and I will update this ad when it does) and feel free to contact any of the group members for informal inquiries.
Position for Postdoctoral Scholar in SQuID
The international working group SQuID (Statistical Quantification of Individual Differences; https://github.com/hallegue/squid) and the Center for Biodiversity Dynamics at the Norwegian University of Science and Technology (https://www.ntnu.edu/cbd) is seeking to hire a POSTDOCTORAL SCHOLAR to join us for 19 months in augmenting three SQuID activities: 1. Developing new biological understanding by blending the statistics of linear mixed models with simulations of biological processes affecting real data; 2. Refining in-person workshops focused on teaching applications of linear mixed models; and 3. Expanding remote or on-line educational materials. We seek a collaborative individual with a PhD in a relevant area and other relevant skills and experience. The preferred candidate should have a strong interest in life science with topical issues in ecology and evolution, have a good knowledge of statistics, be well versed in programming (e.g. R, C++), and possibly have a record of contributing to novel educational approaches to quantitative topics, especially via remote or on-line platforms. The successful candidate would likely start in early 2021, initially be based at NTNU, and would participate fully in several planned SQuID activities, including 6 workshops (currently planned to be held at 6 institutions worldwide) combining teaching and research, and one or more working groups focused on both research and educational initiatives. For further information, you may contact Jonathan Wright (firstname.lastname@example.org), David Westneat (email@example.com), or Niels Dingemanse (firstname.lastname@example.org). An official position description at NTNU and the application procedure will be forthcoming in late summer. Contact one of us to be put on a mailing list for that announcement.
SQuID currently consists 14 scholars interested in the biology of variance and the tools to analyze it. The funded project builds on the core idea that targeted simulations of data under specified conditions can improve analysis of real data and expand biological thinking as a result. This idea and the R package that resulted from initial work on this are described in Allegue et al. (2016). An additional project on violating distribution assumptions is in press (Schielzeth et al. MEE), and projects in the works include a MS on the consequences of mean-centering (in review), the use of diagnostic tools (MS) and temporal biases and possible solutions (MS). The current grant focuses on improving and extending the educational tools in the R package into other venues, and developing new research in 4 new topic areas:
- Multivariate GLMMs combined with structural equation modelling can provide new insights into trait interactions and patterns of covariation in phenotypes linked by pleiotropy and integrated plasticity. The full capabilities of these approaches have yet to be explored for real datasets (with unequal sampling, biases, and missing data).
- Double GLMMs explore patterns in both means and residual variances, which can produce hidden effects and influence the behaviour of conventional GLMMs, and which arise from poorly understood biological processes. However, their biological implications and performance with real data need to be assessed in more detail.
- Animal models are a powerful extension of GLMMs, decomposing variation into genetic versus environmental sources using pedigree information as a random effect. We will test animal model performance with incomplete data and uncertain pedigrees.
- Phylogenetic GLMMs address important evolutionary questions using variation among taxa. Such meta-analyses include information on the sampling design of multiple studies, and issues due to unequal sample sizes, missing information and uncertainty about phylogenies.
The post-doc will be welcome to join any of these research groups, but a critical task assigned to the position is to continue the process of translating new research into educational materials. These are intended for both in-person teaching in workshops and tools for online or remote learning. Ability to program will be a key skill, but knowledge in both biology and education will also be advantageous.
In-person workshops are planned for roughly every 6 months starting in Spring 2021 with a meeting in Trondheim, Norway. Additional workshops are planned for Montpellier (France), Davis (California), Hokkaido (Japan), Sao Paulo (Brazil), and Montreal (Canada).
The current members of SQuID are:
Hassen Allegue: UQAM, Montreal, Canada
Yimen Araya-Ajoy: NTNU, Trondheim, Norway
Anne Charmantier: CNRS, Montpellier, France
Barbara Class: USC, Sunshine Coast, Australia
Eduardo da Silva dose Santos: USP, Sao Paulo, Brazil
Niels Dingemanse: LMU, Munich, Germany
Ned Dochtermann: NDSU, Fargo, ND, USA
Laszlo Garamszegi: IEB, Budapest, Hungary
Kate Laskowski: UC-Davis, Davis, CA, USA
Shinichi Nakagawa: UNSW, Sydney, Australia
Denis Reale: UQAM, Montreal, Canada
Holger Schielzeth: University of Jena, Germany
Celine Teplitsky: CNRS, Montpellier, France
David Westneat: UKY, Lexington, KY, USA
Jon Wright: NTNU, Trondheim, Norway
SQuID Papers (so far!):
Allegue, Hassen, Yimen G. Araya‐Ajoy, Niels J. Dingemanse, Ned A. Dochtermann, László Z. Garamszegi, Shinichi Nakagawa, Denis Réale, Holger Schielzeth, and David F. Westneat. “Statistical Quantification of Individual Differences (SQuID): an educational and statistical tool for understanding multilevel phenotypic data in linear mixed models.” Methods in Ecology and Evolution 8, no. 2 (2017): 257-267.
Schielzeth, Holger, Niels J. Dingemanse, Shinichi Nakagawa, David F. Westneat, Hassen Allegue, Céline Teplitsky, Denis Réale, Ned A. Dochtermann, László Z. Garamszegi, and Yimen G. Araya‐Ajoy. “Robustness of linear mixed‐effects models to violations of distributional assumptions.” Methods in Ecology and Evolution.