August 25, 2025

Authors: Participants in the EFI University for Everyone working group at the EFI2025 Conference, and EFI Education and DEI working group members

Introduction 

The EFI University for Everyone working group activity, held at the EFI2025 Conference (https://ecoforecast.org/efi-2025-conference/), brought together EFI community members from a wide range of career stages and experiences to envision components of a “university education” built on shared foundational values of community and mentorship. In small groups, we identified common themes of place-based and project-based learning to support ecological forecasting education, having fun while learning, and inclusivity, where everyone teaches and everyone learns. The idea was to consider not just the curriculum needed to teach ecological forecasting, but also holistic activities that provide community and support for students, without the constraints of our institutions. We envisioned a curriculum and activities developed for everyone, not only tuition-paying university students. Given that the group had a little over 3 hours over two days at the conference to develop ideas, the foundation was set, but there is more to consider and build on. As you’ll see below, lots of ideas were generated, many of which have relevance in STEM education far beyond ecological forecasting. If any of this is of interest, the EFI Education and DEI working groups are always welcoming new participants. See the Working Group page or the EFI newsletter for the schedule of upcoming calls, or reach out to info@ecoforecast.org for more info. 

After an initial brainstorming session, three main topics were identified to explore further in small groups: curriculum, mentorship, and community engagement. Here we provide a list of action items related to these three topics, and then provide additional detail following that.

The purpose of this blog post is to share the ideas generated with the EFI community, so the EFI Education and DEI working groups have accountability to and from the EFI community, and to provide a starting point or inspiration for future grant applications and activities in EFI.

Action Items

1. Curriculum
   1. Develop tiered standards (e.g. for  high-and low-Internet connectivity conditions and open access material) and identify where the forecasting materials EFI has compiled fit within those tiers
   2. Develop a visual flowchart of the resources
2. Mentorship
   1. Develop and facilitate short-term (one or a series of online EFI community meetings) or long-term (6-10 week ongoing meetings) opportunities for EFI community members to exchange technical and professional knowledge.
3. Community Engagement
   1. Create structure for forecasting education in a math-free (or pre-math) and intuitive manner at the K-12 level
4. Refine the EFI University for Everyone Mascot

Curriculum 

The curriculum discussion was framed around making EFI and ecological forecasting educational material more accessible and inclusive, both in the content itself and the modality of instruction. We found a shared interest in creating new material that is flexible, place-based, and contextualized. As many of the conference working group participants had experience with creating and sharing open educational resources, we asked ourselves what assumptions exist (such as internet connection, technology, etc.) and who are we leaving out when we make these assumptions? To make our community of environmental data science and ecological forecasting more inclusive, we aspire to develop tiered standards for the EFI educational material that range from offline compatibility to in-person courses, across different tools and modalities, with a focus on open access resources (similar to FAIR and CARE principles). We plan to inventory and organize the existing list of ecological forecasting online educational resources (Willson et al. 2021; https://doi.org/10.1002/ece3.10001) into these standards (Table 1), and develop a visual flowchart (similar to Figure 1 in Lofton et al. 2024 in BioScience; https://doi.org/10.1093/biosci/biae089). 

We acknowledge that individuals learning about ecological forecasting come from different pathways, and learning doesn’t have to be linear. To advocate for education at all levels, whether enrolled in a university course or not, we want to create resources for learners to develop skills to do research in ecological forecasting at their own pace and create their own pathways, similar to the braided river analogy for STEM workforce development in Figure 1 below (Batchelor et al. 2021; https://eos.org/opinions/reimagining-stem-workforce-development-as-a-braided-river). 

Figure 1. Braided river example of developing an inclusive STEM workforce (from Batchelor et al. 2021)

One example that was shared to address challenges related to teaching computational approaches in internet-limited environments came from the Carpentries Offline community, which has been developing microcomputer (e.g., Raspberry Pi) based tooling that allows teachers to provide learners access to common tools for teaching and doing data science through a local Wi-Fi connection that does not require internet access. This infrastructure is primarily designed for teaching workshops, but the group discussed potential applications for individual learning in internet-limited environments (e.g., doing homework off campus, for courses in prisons or in remote areas).

One actionable item is to evaluate available ecological forecasting curricula in terms of their internet accessibility, along with the learning style (self-directed to classroom settings). We recommend the following matrix to evaluate the suitability of ecological forecasting curricula (Table 1). Even though we have discrete categories, access and learning style fall along a continuum. We have a shared community valuation of open-access educational materials for ecological forecasting. 

Table 1. Matrix of tiered standards for the EFI educational material that range from offline compatibility to in-person courses, across different tools and modalities

Accessibility ↓ / Learning style →	👤Self-directed  (i.e., learning on your own)	👥Group (i.e, as part of a learning community; no direct assessment).	🏫 Classroom (i.e., includes a teacher and student and independent assessment of knowledge)
Internet limited; software limited (i.e. intermittent internet access or limited to hotspot usage; bundled software as part of a computer / smart phone installation)
Internet access; software limited (i.e. access to open source tools such as R, Python; internet access reliable enough to access Google products)
Internet access; software access  (i.e. access to reliable, high-speed internet, fee-based cloud computing services (Google Earth Engine; AWS, and others GEE), and proprietary processing software

Mentorship

The mentorship discussion stemmed from the unifying principle of knowledge exchange: everyone serves simultaneously as a teacher and a learner, without strict adherence to traditional hierarchies of mentorship. The group discussed specific ideas of how to implement mentoring activities in EFI. 

Each participant in a mentoring activity would join as part of a cohort, which partakes in mentorship in one or more modes: through one or a series of online community meetings (short term), or through a 6-10 week program (long term). This may be best situated by getting involved with, and possibly building upon, existing related initiatives, such as the BioGraphI undergraduate teaching mentorship network or applying the structure originally used for networking by the Center for Scientific Collaboration and Community Engagement (CSCCE; https://zenodo.org/records/15643151). 

With both modes, and prior to engaging in network activities, each applicant would provide information on their background, interests, and goals. This could occur either by filling out a survey or via informational interviews conducted by organizers. Efforts would be made by organizers to encourage applicants to view and communicate themselves, and others, holistically beyond professional titles, inclusive of their familial, community, and spiritual identities, as some examples. Organizers would then compile such applicant information to designate pairs or triplets with shared or complementary identities. 

Such small groups would meet individually to exchange knowledge, which likely would include technical and professional knowledge that is directly pertinent to ecological forecasting but could also include more broadly defined interests and skills, including natural history, music, and art. Especially in a long-form mode, pairs/triplets would exchange and circulate with different members, to maintain targeted exchange between individuals while also enhancing cohort-wide communication. 

We also discussed the importance of identifying incentives for recruitment, which may include learning new skill sets, external validation of one’s current abilities, and network building. This fluid model of partnership and co-mentoring could be understood to resemble the dynamics of the square dancing that EFI2025 Conference participants so enthusiastically partook in during our evening at Mountain Lake Lodge – folks are paired up as part of a group of dancers, and learn alongside each other while switching partner and group membership to create diffuse rather than top-down opportunities for connection. Like a square dance caller, mentorship network organizers coordinate opportunities for participants to interact, but allow participants to decide for themselves the nature and tone of interaction.

Figure 2. EFI community members square dancing at the EFI2025 Conference at Mountain Lake Lodge in May 2025

Community Engagement

The discussion around community engagement and connections stemmed from the conference working group participants’ personal experiences with organising workshops at various levels and geographies. At least two main interests emerged: 

1. Forecasting education in a math-free (or pre-math) and intuitive manner and 
2. Working at locations/contexts where community engagement is ongoing and long-term, and introducing forecasting through collaborations. 

Several forecasting resources already exist, but our unconfirmed sense was that these tended to be more technical and targeted at skill building rather than intuition-building, and at undergraduate or later levels rather than K-12. From an early age, science is often taught in the hypothetico-deductive method, which has a lot in common with the iterative learning process of ecological forecasting. So it would be possible to transition scientific education to incorporate ecological forecasting by: 

1. Developing outreach and easy-to-use curricular material to help K-12 instructors who are unfamiliar with ecological forecasting
2. Providing simple success stories about how to incorporate uncertainty as motivators for a concept that can be overwhelming

Forecasting really shines in an iterative setting, which is a real challenge for educational settings that are short-term, or where resources aren’t available for persistent exercises, or where students are in-and-out of the educational setting. Teaching on ecological forecasting in a K-12 setting can provide a stimulating gateway by which to engage learners with science and math. Ecological applications can connect with charismatic concepts and agents that may resonate with young learners (e.g. wildlife or sunny/rainy weather) and can also provide an avenue to confront ideas and predictions with new observations in real time (i.e. near-term forecasting that can soon be confronted with reality).

The conference working group participants developed these ideas as an example for engaging with the community with an intuition-based teaching module for forecasting with middle school students in a summer camp as the target audience, as a summer camp is free of the constraints of a regular school, such as state and district evaluations. The module could be taught over six weeks with three 4-hour days in a week. There would be five weeks of activities with a final week for presentations of tangible outcomes at camp and/or at a public event, meeting, or conference.  

The weeks would cover Data, Uncertainty, Relationships, Models, Predictions. These would be taught using simple graphical tools such as paper/pen drawings of x-y relationships. Learners could be asked to hand-draw the relationship they expect and a forecast. Learners would go through these processes by observing something outside, although we did consider thinking about potential observations that could be done inside a classroom, e.g., a kit that can be sent to a classroom interested in teaching forecasting (à la Srikant and Aggarwal 2017; https://dl.acm.org/doi/10.1145/3017680.3017717). 

There are two threads that would be woven through the activities. The first would be: I see, I think, I wonder – as jargon-free ways of explaining Data, Relationships, and Predictions. Second would be using a hypothesis as a mental model of the world. These could be derived from any worldview, and forecasts could be built using multiple hypotheses, thereby allowing different ways of knowing to be incorporated into the module.

The ideas for a summer camp were inspired by Dalbotten et al. (2014; Journal of Geoscience Education; https://doi.org/10.5408/12-408.1)

Proposed Mascot for EFI University for Everyone 

The group had fun developing potential options for a mascot, because every university needs a mascot!

Figure 3.  Initial groundhog mascot for EFI University for Everyone. Notice the groundhog’s shadow in the background.

Figure 4. Alternative groundhog mascot for EFI University for Everyone