Overview of the Macrosystems EDDIE curriculum

The Macrosystems EDDIE ecological forecasting curriculum for undergraduates comprises four standalone modules: Introduction to Ecological Forecasting, Understanding Uncertainty in Ecological Forecasts, Using Data to Improve Ecological Forecasts, and Using Ecological Forecasts to Guide Decision-Making (Fig. 1). Like all EDDIE modules, Macrosystems EDDIE ecological forecasting modules are designed using the 5E (Engagement, Exploration, Explanation, Expansion, Evaluation) instructional model (Bybee et al. 2006), which is implemented through a scaffolded A-B-C structure (O’Connell et al. 2024). In all modules, Activity A Engages students and asks them to Explore the module’s focal topic, Activity B further Explains and asks students to Expand on that topic, and Activity C Evaluates students’ understanding of the topic (Carey et al. 2015, O’Reilly et al. 2017). The three-part scaffolded structure also maximizes the adaptability of Macrosystems EDDIE modules to various classroom contexts, as instructors can choose whether to complete just Activity A, Activities A and B, or all three activities in one to three-hour course periods. Each module can be taught individually or instructors may choose to implement multiple modules throughout their curriculum; example use cases are detailed in the Course implementation section below.
The modules in the Macrosystems EDDIE ecological forecasting curriculum are designed to both 1) introduce ecological forecasting concepts and 2) develop data science skills (Fig. 1). To accomplish the first goal, each module covers a foundational concept in ecological forecasting, and students then apply the forecasting concept to a NEON lake site of their choice. To develop data science skills, students use environmental data collected by NEON (Keller et al. 2008, Goodman et al. 2015) as the basis for their forecasting analyses. Working with NEON datasets requires students to evaluate the quality of the data (e.g., gaps, outliers, biases) and confront how inherent variability and error in environmental datasets may affect their analyses. In addition, each module asks students to interpret data visualized using various methods, ranging from time series and scatterplots to probabilistic forecasts and histograms. Finally, each module focuses on one or more foundational quantitative skills in ecological forecasting, including building and calibrating ecological models, generating forecasts, quantifying the uncertainty associated with predictions, using new observations to update forecast models, and designing forecast visualizations to effectively communicate forecast output.
Macrosystems EDDIE modules include a comprehensive set of instruction materials and are suitable for implementation in a variety of class contexts (Fig. 1). All modules are delivered through an R Shiny interface, where R code is used to render a website that students can access in their internet browser (Chang et al. 2023). This permits a user-friendly, point-and-click interface for introductory students and aims to lower the intimidation barrier to ecological forecasting, as students do not need to have any coding skills to generate a forecast. For classrooms where gaining R coding skills is a learning objective, two of the modules (Understanding Uncertainty in Ecological Forecasts and Using Data to Improve Ecological Forecasts) have Rmarkdown activities in addition to R Shiny materials. The Rmarkdown activities enable students to access and modify the code underlying the R Shiny app and complete module activities in the R programming environment (Xie et al. 2018).
All Macrosystems EDDIE ecological forecasting materials are designed to provide instructors with “just-in-time” training (sensu Novak et al. 1999) on data science skills as they prepare to teach the modules in their classrooms. In addition to the R Shiny application (and RMarkdown file if applicable), each module includes an introductory (~30 minute) Microsoft PowerPoint presentation with slide notes; a Microsoft Word student handout with pre-class readings, activities, and questions associated with the module; a comprehensive instructor manual with learning objectives; detailed guidelines for module implementation and answer keys; and a “quick start” guide to the R Shiny applications. Notably, instructor manuals include strategies for teaching and recommendations for implementing the modules across a variety of course schedules (e.g., three, one-hour class sessions vs. one, three-hour lab period) and modalities (e.g., virtual, face-to-face, hybrid).
All module teaching materials are licensed under the CC BY-NC-SA 3.0 license allowing modification for classroom use and are published in the Environmental Data Initiative repository (Moore et al. 2023a, 2024b, Woelmer et al. 2023b, Lofton et al. 2024c), and all module code is published in the Zenodo repository (Moore et al. 2023b, 2023c, 2024a, Woelmer et al. 2022, Lofton et al. 2024a, 2024b). In addition, all module code is maintained and updated at the Macrosystems EDDIE GitHub organization (https://github.com/MacrosystemsEDDIE). We encourage and welcome instructors and students to adapt and modify these materials for their classrooms, projects, and research.