This RAPID Design and Development Learning Strand project used the science of hurricane resilience to foster the development of student agency and appreciation for science in Puerto Rico. For school age children on the island of Puerto Rico, the back-to-back hurricane strikes of Irma and Maria in 2017 were their first experience with the tragedy of hurricanes in Puerto Rico. There was much concern in the media and associated reader comments about the ability of the forests in Puerto Rico to recover. These concerns reveal common misconceptions about the dynamics of forest ecosystems. Long-term research conducted by Luquillo ecologists paints a different picture. The rainforest is adapted to being struck by hurricanes on a regular basis. Engaging in investigations of El Yunque resilience at different timescales and different perspectives allows students to connect their learning about the rainforest to the current tragedy of their everyday life. 

The Luquillo Schoolyard program supported two strands of activities: Journey to El Yunque and Data Jam. With Journey to El Yunque, middle school students study past hurricane disturbance and resilience. With Data Jam, high school students investigate the hurricanes in the context of past storm events using long-term data from the Luquillo LTER. In both cases, students gain a scientific understanding of the events that have occurred to them and of the resilience of the rainforest. In addition, these investigations provide a positive outlet for students to express their feelings in creative ways. Researchers hypothesized that the quality of implementation would be predictive of the quality of students' projects, increased understanding of the dynamics of resilience in El Yunque, as well as increased appreciation for science.

Journey to El Yunque

During the 2018-19 school year, ten teachers in Puerto Rico participated in professional development and implemented the program with 183 students. Over 90% of the students were Hispanic and two-thirds of the students were female. Students were administered a pre and post standard-based assessment of ecology as well as a pre and post attitudinal survey. The assessment items were fit to a Rasch model and scaled to a 10-point scale. Overall, students statistically increased their performance from pretest (4.5/10) to posttest (4.9/10) with an effect size of 0.22 (p<0.01).  As an indicator of fidelity of implementation, we surveyed students about the extent to which they engaged in a variety of inquiry-based activities during the unit. The survey was adapted from the Inquiry-Based Science Instruction scale (Allensworth et al., 2008).  Since teachers have multiple students, we used multilevel measurement modeling to develop an index score for each teacher that aggregates scores from multiple students by teacher. We conducted the multilevel measurement modeling using WHLM software version 7.24q with student demographic and assessment scores at level 1 and the aggregate Inquiry-Based Science Instruction Score at level 2 for each teacher. At the student level, after controlling for pretest performance, there was no difference in performance by gender. At the teacher level, the Inquiry-Based Instruction score was a statistical significant predictor of the posttest performance. In other words, these results provide evidence for the hypothesis that the quality of implementation predicted increased understanding of population dynamics.

Data Jam

In Data Jam, students investigate their own research question using long-term data from the Luquillo LTER. Each year, the teachers select a subset of projects to present at the annual Luquillo Schoolyard LTER program. We collected the final projects for students who presented at the Luquillo Schoolyard Symposium. For this project, we scored and analyzed the projects from the first year of the grant and from the last year of the grant. The rubric derives from research on reasoning with data, especially using interactive visualization tools.  It focuses on five dimensions: (1) appropriateness and complexity of research questions, (2) distributional thinking and awareness of variability, (3) attention to both aggregate values and individual cases, (4) data moves, (5) communicating research findings.  We scored 10 posters from the 2017-2018 Luquillo Data jam and 7 posters from the 2020-2021 Luquillo Data Jam.  For each rubric attribute, the posters were scored on a scale of 0 (low) to 4 (high). All five attributes increased from 2018 to 2021. Four of the five increases were statistically significant. The only dimension that was not a statistically significant increase was communication of findings. While there is much to celebrate in the significant increases, there is still more work to do. None of the attributes averaged above 2 points (medium) on a four-point scale. 

Allensworth, E. A., Correa, M., & Ponisciak, S. (2008). From High School to the Future:  ACT Preparation?Too Much, Too Late Chicago: Consortium on Chicago School Research at the University of Chicago.

Last Modified: 08/09/2021
Modified by: Steven M Mcgee