Analyzing Trajectories for Learning About Ecosystems

• In Event: Developing Evidence of Understanding Complex Systems

Abstract

Learning research has documented the challenges of learning about complex systems (e.g., Jacobson 2001) and has demonstrated what students learn after participating in systems instruction (e.g., Hogan & Fisherkeller, 1996). Yet, few studies have documented the learning trajectories that students take to understanding ecosystems. To do so, systems thinking must be characterized by tasks that elicit complex performance and that can be administered multiple times during an instructional intervention. This poster describes our approach to these challenges and explores how middle school students may develop an increasingly coherent understanding of ecosystems.
The context for this analysis was a six-week enactment of a new curriculum unit designed to promote systems thinking and student understanding of aquatic ecosystems. To facilitate student thinking about multiple interacting system components, we designed computer-based tools and curriculum around the structure, behavior, and function (SBF) conceptual representation (Hmelo-Silver et al, 2007). An SBF model of a system depicts its [S]tructures (configuration of components and connections), [F]unctions (output), and [B]ehaviors] (internal causal processes and mechanisms that enable components’ functions) In this way, SBF guides knowledge organization by focusing attention on relations, thus enabling learners to reason about multiple interacting components and functions and the underlying causal mechanisms.
We build upon an exploratory study (Eberbach et al, 2012) using a larger sample and new anlyses. A microgenetic analysis used data sources that included 895 student drawings from two sets of drawing assessment tasks. The first set, which was administered before and after classroom instruction, asked 229 students to draw what happens in an aquatic ecosystem. The second set, the Aquarium Assessment (AA), was administered twice during classroom instruction. Based upon what they learned in class, students were asked how they would set up an aquarium to keep fish healthy. We coded and scored each drawing at three levels (isolated (1) to integrated (3), along multiple dimensions (Biotic/Abiotic, Macro/Micro, SBF, Extraneous Structures) to detect how interacting components and processes may affect increasingly complex systems thinking (Table 1).

The following graph depicts the group means for each systems dimension over time. A visual inspection reveals some notable patterns. First, each systems dimension developed along different trajectories, as evident by the variation in their slopes. Second, growth of the SBF slope appears to correspond with the timepoint at which students also included relations between Macro/Micro levels and between Biotic/Abiotic levels. Third, these three dimensions appear to approach convergence by posttest. Furthermore, many students associated system structures with behaviors or functions from the start. However, these connections largely occurred at relatively simple biotic and macroscopic levels. Finally, students included few extraneous features at pre-test and their frequency diminished over time.

Figure 1. Systems dimensions developed along different trajectories

Using a microgenetic approach with an easy-to-administer task made it possible to better understand the complex nature of what it may mean to learn about ecosystems in a classroom. This is a useful reminder that learning does not necessarily occur in linear patterns and that different dimensions of learning may follow different paths (Wilson, 2009).

Authors

• Catherine Eberbach, National Science Foundation

• Cindy E. Hmelo-Silver, Indiana University

• Rebecca Jordan, Rutgers University