Middle

Large language models (LLMs) have demonstrated strong potential in performing automatic scoring for constructed response assessments. While constructed responses graded by humans are usually based on given grading rubrics, the methods by which LLMs assign scores remain largely unclear. It is also uncertain how closely AI’s scoring process mirrors that of humans or if it adheres to the same grading criteria. To address this gap, this paper uncovers the grading rubrics that LLMs used to score students’ written responses to science tasks and their alignment with human scores. We also examine whether enhancing the alignments can improve scoring accuracy.

Large language models (LLMs) have demonstrated strong potential in performing automatic scoring for constructed response assessments. While constructed responses graded by humans are usually based on given grading rubrics, the methods by which LLMs assign scores remain largely unclear. It is also uncertain how closely AI’s scoring process mirrors that of humans or if it adheres to the same grading criteria. To address this gap, this paper uncovers the grading rubrics that LLMs used to score students’ written responses to science tasks and their alignment with human scores. We also examine whether enhancing the alignments can improve scoring accuracy.

The disproportionate impacts of societal challenges (e.g., climate change, air and water pollution) on minoritized groups expose systemic injustices and compels STEM educators to reframe the role of STEM education in society. In this article, we describe traditional approaches, contemporary approaches, and our proposed future approach in science and STEM education with a focus on equity and justice.

Prior research has explored innovative ways to introduce data science to young learners, emphasizing not only the development of data skills but also the connection of data science to students’ authentic inquiries and critical actions. Building on this foundation, this study aims to achieve two complementary goals: integrating Knowledge Building, a well-established pedagogical approach, into K-12 data science education, and enhancing students’ epistemic agency through data practices in knowledge building.

The Language, Culture, and Knowledge-building through Science (LaCuKnoS) project tests and refines a model of science teaching and learning that brings together current research on the role of language in science communication, the role of cultural and community connections in science engagement, and the ways people apply science knowledge to their daily decision making. One key component of the model brings families together as co-learners and co-teachers through family learning experiences. We describe our work to promote more robust family conversations about science in our lives within an existing research practice partnership, using a two-tiered qualitative conversational analysis to compare the family conversations that result from three family engagement models: (a) family science festivals; (b) family science workshops; and (c) family science home learning.

The Exploring Coasts and Coastal Change in Hawai‘i unit supports middle school haumana (students) in developing multi-perspective understanding and personal stances about coastal change in their community. The unit was collaboratively developed by a partnership among educators bringing together Indigenous and Western ways of knowing and learning. Through lessons that take place over three weeks, haumana (students) undertake a place-based learning experience addressing physical, biological, and social facets of their changing coasts and engage the perspectives of ‘ohana (family), community members, and scientists.

As educators, we recognize that commercially prepared curricula advertised as “NGSS aligned” do not necessarily emphasize student sensemaking. In this article, we describe our process of modifying such curricula by reflecting on previous instruction and planning for future instruction that centers student sensemaking in a middle school unit on chemical reactions.

This study practically addresses some key challenges teachers face in enacting reform-oriented science teaching and offers suggestions for how continued research regarding norms and uncertainty can continue to further science reform efforts.

We present findings from an analysis of tests of teacher mathematical knowledge identified over a 20-year period of mathematics education literature, and report on how these tests are discussed in the literature, with a focus on validity arguments and evidence.