Over the lifetime of our NSF award titled "EAGER: Managing our expectations: quantifying and characterizing misleading trajectories in ecological processes" our collaborative research group has developed and refined a "broken window" algorithm. This algorithm has been used to analyze long-term datasets in order to better understand how much data is needed to more accurately forecast ecological trends in studied communities as well as identify misleading trends in existing long-term data sets. The results of this study will guide how data, particularly processes that unfold over time, are interpreted in ecology. It will also help to prioritize which research directions require long-term study and which can be understood at shorter time scales, an important aspect of completing research effectively and efficiently. Through papers, preprints, conference talks and posters, and blog posts, we have been able to disseminate findings to a broad audience of scientists and learners, an important aspect of science communication.

We have written 10 blog posts, five of these led by undergraduate trainees, and we have written and published a Data Nugget, titled "Blinking Out?", for a 10-12th grade student audience discussing the importance of long-term data using fireflies in southwest Michigan as our model population. This Data Nugget will be used in classroom activities designed to familiarize students with understanding scientific concepts, graphing and analysis, and using data to identify trends and make predictions. 

While identifying appropriate datasets to apply the algorithm to, we were able to provide immersive training opportunities to five undergraduate students that has resulted in five blog posts that document the important work they?ve completed, as well as their learning experiences, for the overall project. Writing the blog posts on their experiences not only helped undergraduate trainees identify obstacles and issues that come with identifying, cleaning, and reusing existing long-term datasets in an ethical manner, but honed their narrative writing skills for a broad audience that will serve them well in future careers in environmental data science and management.

This project has allowed for in-person and virtual travel to present findings at eight local, regional and national conferences, which not only aids in disseminating research findings but also allows for professional development for senior and undergraduate team members. This is especially important for undergraduate students because it helps them gain confidence in presenting to a wide audience while honing their presentation skills and understanding of how research is shared while also fostering future research collaborations and possible graduate opportunities.

Our collaborations with research scientists from external organizations have helped us test, refine, and apply the algorithm to current datasets, leading to a broader understanding of the algorithms applications, limitations, interpretations, and scientific impacts, which has resulted in scientific publications in the journals Global Change Biology, Ecological Informatics, and Ecology Letters, journals that reach a broad scientific audience.

 Accessibility and equity in science and data reuse are paramount to this entire endeavor, and we have written seven blog posts on the subjects of the algorithm, data reuse, and equity in data sharing. Additionally, project funds and resources were used to support eight trainees working at the intersection of ecology and data science: Senior person Perrone and five undergraduates, and two graduate students participated in peripheral projects. Seven of the eight individuals are women and/or underrepresented groups in these fields including one woman of color, three individuals of nonbinary gender, and one person with a disability, and we are providing career-stage appropriate mentoring, guidance, and opportunities to help support them as they pursue careers in environmental data science.

Last Modified: 12/13/2021
Modified by: Christine Bahlai