Course outline


Loading wiki pages...

Wiki Version:
<h3>Course outline</h3> <h4>Leaders: Laura Pollock (McGill) and Jennifer Sunday (McGill)</h4> <p>Synthesis statistics for ecology and evolution This course will provide an introduction and overview of approaches for synthesizing the highly structured, multi-sourced datasets that typify ecology, evolution, and environmental research, including data collation, integration, analysis, and visualization. Students will learn how to define their research question and scope in the context of other studies in their subfield and how to find and fill in gaps in understanding with additional datasets. We will also provide a broad overview of the different types of methods used for synthesizing data including hierarchical models, meta-analysis, model integration, and model updating, providing students a guide to navigating these methods and identifying methods to learn and use in their research. Students will develop skills in R programming, collaborative research, reproducible workflows, data analysis, and communication.</p> <h3>Delivery format</h3> <ul> <li>8 sessions, 1.5 hours per session</li> <li>each session generally includes a lecture and hands-on component, each varying in length among sessions</li> </ul> <h3>Student assessment</h3> <ul> <li>Session assignments (5) : 20% each</li> </ul> <h3>Required materials</h3> <ul> <li>Personal computer (laptop or desktop)</li> <li>internet connection</li> </ul> <h3>Pre-requisites</h3> <ul> <li>Graduate-level thesis in Ecology or Evolutionary Biology</li> <li>Introductory coding experience or a willingness to learn</li> </ul> <h3>Tentative schedule</h3> <p>Class 1. Overview of the course and expectations. The benefits of synthesis in ecology. In this class, we will discuss the reproducibility crisis in ecology and science more generally, and the need for putting individual studies in the context of the entire subfield.</p> <p>Class 2. How to go from a research question to a model. How to ask questions like a biologist and frame them like a statistician. Forming questions and expectations What are the response and predictor variables? What is the process? Writing out your mental model (boxes and arrows)</p> <ul> <li>Assignment 1: Write out a mental model for your research question. This ‘model’ will serve as the basis for subsequent classes. In break-out groups, describe your model to peers for peer feedback. Submit your final mental model.</li> </ul> <p>Class 3. An overview of data synthesis and meta-analysis. Understanding the differences between synthesizing already analyzed datasets (formal meta-analysis) and synthesizing multiple datasets (data integration). </p> <ul> <li>Assignment 2: Use your research question/mental model to develop plans for a project using a meta-analysis and/or data integration approach. Research what data exist to address this question, either in many individual published records or in databases. Summarize your question and the state of the data.</li> </ul> <p>Class 4. Securing and cleaning datasets Where to get data to supply the variables of your model, what types of data are available for data synthesis and meta-analysis How to assess coverage of datasets Cleaning data using reproducible workflows Matching multiple datasets (e.g. taxonomy, location data) Interpolation</p> <ul> <li>Assignment 3: Working with an assigned dataset or your own, identify errors and gaps in the data with exploratory figures, correct errors through cleaning scripts. Express the pros and cons of data interpolation.</li> </ul> <p>Class 5. Introduction to meta-analysis How to search the literature Reported statistics &lt;-&gt; research question Variation among vs within datasets Common biases (publication, non-independence, etc.)</p> <p>Class 6. Meta-analysis II Meta-analysis vs. extracted data synthesis Bias vs. noise in extracted data Accounting for sources of variation - Assignment 4: Find a formal meta-analysis or extracted data synthesis in your field and evaluate it using the concepts presented.</p> <p>Class 7. Introduction to Data Synthesis Statistics - Overview of common methods for synthesizing large datasets, including hierarchical models that can be used to depict with- and between-site variation or different taxonomic levels, and formal model integration where different types of input data can be integrated.</p> <p>Class 8. Data Synthesis Statistics II - We will discuss the concept of using prior information combined with newly collected data to update models, reduce uncertainty and refine predictions.</p> <ul> <li>Assignment 5: Find an example of a data synthesis in your field using one of the techniques discussed in Class 7 or 8. Using this example, apply a similar framework to your research question. Option 1: refine your mental model to create a more specific model that could be fit using one of these techniques. Option 2: explore software to test one of the example models with your data or example data and briefly report your results.</li> </ul>
OSF does not support the use of Internet Explorer. For optimal performance, please switch to another browser.
This website relies on cookies to help provide a better user experience. By clicking Accept or continuing to use the site, you agree. For more information, see our Privacy Policy and information on cookie use.

Start managing your projects on the OSF today.

Free and easy to use, the Open Science Framework supports the entire research lifecycle: planning, execution, reporting, archiving, and discovery.