SimScientists: Use of Learning Progressions Within Curriculum-Embedded Simulations for Formative Assessment

• In Event: Approaches to the Use of Learning Progressions Within Multidimensional Formative Assessment Activities in Science

Abstract

1. Purpose: This presentation will describe the development and validation of a suite of curriculum-embedded science simulations for formative assessment of core concepts and inquiry practices, including building and using system models, and that can produce individualized, graduated coaching and progress reports to the teacher to support differentiated instruction. The rationale for use of simulation-based science assessments, their grounding in theory and research, their use of learning progressions, and the evidence from pilot and field testing in two states of the technical quality of the assessments, feasibility of their implementation across a broad range of contexts, and impacts on student learning will be shared.

2. Theoretical Frameworks: Conceptual, physical, and computation modeling, called for within the NGSS, are practices by which scientists create and communicate understanding of science systems (Nersessian, 2008). Simulation-based models of science phenomena can permit investigations of causal, temporal, and spatial relationships (Gobert & Clement, 1999). A growing body of research shows that once learners internalize schema-based models of complex system organization—components, interactions, and emergent behaviors—learners can transfer this heuristic understanding across systems (e.g., Goldstone, 2006; Goldstone & Wilensky, 2008).

Over the past decade, learning progressions have moved to the center of discussions about the improvement of formative and summative assessment. However, we have learned that learning progressions are neither developmentally inevitable nor universally applicable – that there is a complex interaction of instruction and prior knowledge in the development of new knowledge (Corcoran and Mosher, 2009; Shavelson, 2009).

3. Methods: In the design and development of this suite of simulations to support the development of deeper understandings of a system model of atoms, molecules, the principles of Evidence Centered Design and Universal Design for Learning were applied. A system target model laid out the science knowledge and practices to be tested and organized the concepts and inquiry practices to be assessed into the three system levels of components, interactions, and emergent properties of matter. These were used to design a sequence of tasks that assessed both core ideas and science practices (the student model). Within these sequences, task models were designed to shape activities intended to elicit observable evidence of student knowledge and use of practices and individual progress in core understandings.

4. Data Sources: Evidence of the technical quality, feasibility, and usability of the simulation-based formative assessments was gathered from a number of qualitative and quantitative sources throughout the project. Expert judgment, cognitive labs, and pilot testing were utilized, followed a small randomized control trial comparing students in classes that took the embedded, formative assessments with classes that did not take them (Quelmalz et al, 2013).

5. Results/Substantiated conclusions: The above mentioned studies concluded:
• The embedded assessments had significant impacts on learning.
• The assessments had high technical quality and could be implemented widely.

6. Significance: This particular formative assessment of complex competencies in science provides insight into how learning progressions might be used more broadly in the development of formative assessments that lead to improvements in student learning.

Author

• Edys S. Quellmalz, WestEd