A challenge for environmental and ecology education is to convey systems understanding, including how multiple interacting factors can generate multiple possible outcomes, given uncertainty associated with our knowledge of these systems. Qualitative reasoning (QR), an artificial intelligence modeling technique, provides one solution because it allows explicit representation of complex conceptual and causal relations between objects, ideas, and quantities even when numerical information is lacking. Models describing a variety of environmental sustainability scenarios have been developed using QR, but learners need support to explore models so they can construct understanding of issues being represented. Hence, we developed a lesson format designed to support learning from any QR model. The lesson includes a set of ten questions and instructions that guide learners through four categories of model content: system structure, causal relationships, dynamic consequences, and application and evaluation to real-life situations. We conducted a two-stage evaluation of the progressive learning route with university botany students. The first evaluation was an analysis of the ability of students to comprehend model representations and provide detailed, well supported, written responses to questions in the lesson, based on a model addressing stakeholder influences on sustainability decision making. Questions and diagrams were presented to the class with a video projector. The second evaluation assessed student understanding of plant-resource interactions by comparing differences in pre- and post-test scores (same questions) between experimental and control groups of students. Control students watched a movie related to course objectives but on a different topic while experimental students were provided with printed questions and instructions on how to use the modeling software to generate necessary model results.

Results/Conclusions

Our initial evaluation demonstrated that students could understand model representations and provide thoughtful responses based on them. Students self-reported to be interested and engaged by working with the models, and their responses showed a high level of understanding overall. Many students were able to successfully reason about multiple possible outcomes and complex feedback cycles based on model representations.  In the second evaluation, students improved their test scores after completing the lesson (P = 0.018 and P = 0.017 for two groups of students) but not after completing the alternative activity (P = 0.30 and P = 0.50).  These results indicate that the progressive learning route we developed can facilitate learning content in environmental sustainability and ecology domains using QR technology and shows promise for learning especially in dispersed settings.