Puffins!: Scientific Sensemaking to Improve Middle School Students’ Attitudes Towards Data

• In Event: How Data Science Education Can Enable Exploration of Real World Phenomena

Abstract

Background
The study of puffins among Maine middle-school science students provides a fascinating subject for integrating ecology, ornithology, and data practices. Atlantic puffins have made a remarkable comeback in Maine over the last 50 years, under the leadership of ornithologist Stephen Kress and the National Audubon Society (Kress & Jackson, 2020). Studying the extent to which this seabird population has changed, and the reasons for the changes, is the focus of the Puffins curriculum project (Tumblehome, n.d.; NSF DRL-2241777). The curriculum combines a place-based narrative with data activities, work with AI, and hands-on experiences, such as building puffin burrows. The curriculum employs the integrated approach of scientific sensemaking (Cannady et al, 2019; Mokros et al., 2025) in which scientific knowledge is inseparable from scientific methods like analyzing data.

Objectives
The project aims to: 1) Determine how students use and interpret patterns and relationships in Audubon’s historic datasets about puffins; 2) Examine how students engage in scientific sensemaking as they learn about the data and AI tools that scientists use, and; 3) Show how students’ attitudes towards data (“Data Dispositions”) change as a result of their participation in this three-week curriculum. Our poster will highlight findings about students’ Data Dispositions, which we hypothesize reflect how students engaged in scientific sensemaking experiences exploring authentic data about puffins in Maine.

Methods
The Data Dispositions survey, developed by the Data Clubs project (2021) based on work by the Activation Lab (2016), is a 19-item survey we are using to assess changes in three areas: Fascination with data, self-reported data competencies, and values about data. We are using a retrospective pre-post design where students take the survey only at the end of the unit, recalling their attitudes towards data before the unit as well as their current attitudes. Little et al. (2019) have shown this design is a valid alternative to a traditional pre/post design for dispositional surveys for students as young as fourth grade; students are also less likely to overestimate their confidence before the intervention when reflecting back in time.

Results
Results are presented for 45 students who piloted the retrospective pre-post Data Dispositions assessment in spring 2025. We found that for all three scales (Fascination, Competencies, and Values), median scores increased with statistical significance (p < 0.05) and high statistical power (>0.8), with large effect sizes (>0.5) for Competencies and Values. Work is ongoing, and we will report on results from 500+ additional students participating in fall 2025.

Significance
Our work shows that data science is an ideal venue for scientific sensemaking, because in order to understand the phenomenon being studied, students must explore the data. Students are in a better position to make sense of the data if they care about the scientific questions and feel competent in addressing them. In other words, students benefit from having and developing positive data dispositions.

Authors

• Jacob Sagrans, Tumblehome, Inc.

• Janice R. Mokros

• Pendred E. Noyce, Noyce Foundation