Tammy Long, Michigan State University;
Elena Bray Speth, St. Louis University;
Jennifer Momsen, Michigan State University
Undergraduate introductory biology courses traditionally focus on memorization of linear sequences of facts and concepts. Chapters in texts imply discrete boundaries between one topic and the next and provide useful endpoints for assessing learning. Multiple-choice tests prevail as a means for assessing learning, particularly where resources are scarce but students are not. The image we conjure is not a flattering one. Not only does it promote a view that fragments the discipline into small bits of knowledge, it does not reflect the way in which biologists work. Biology is, in fact, the study of complex living systems, characterized by emergent properties, multiple hierarchical levels of organization, networks of interactions and feedback loops. Increasingly, biological questions are intersecting with other disciplines and biologists are incorporating the language and tools of modeling, computation and engineering to communicate their science.
Our work with complex systems begins in the classroom. As introductory biology instructors, we depart from the linear model of instruction and investigate ways to engage students in learning that will help them develop the skills required to manage biological complexity. To demonstrate understanding of biological systems, students must be able to identify system components and the interconnections among them, predict sources and consequences of feedbacks, and resolve apparent discrepancies between emergent properties of whole systems versus properties of constituent parts. Instructional models that focus exclusively on content and rely on students' capacity for recall will not advance our goals of helping students manage complexity, nor facilitate their understanding of how scientific knowledge is constructed and applied.
We are exploring the extent to which an introductory biology curriculum infused with tools and practices that reflect disciplinary epistemology can improve students' ability to manage complexity in biology. We are specifically interested in the potential of using student-constructed conceptual models and structured arguments to reveal student thinking and promote metacognitive skills.
Models and arguments are foundational tools in biology. Creating and evaluating models of biological systems engage higher-level thinking skills, such as identifying relevant concepts and proposing meaningful connections among them, interpreting relationships within models, discerning a model's purpose, evaluating completeness and accuracy of information represented, and predicting consequences of model perturbations. Structuring a scientific argument requires using data, or evidence, as the basis for constructing a concise and coherent claim supported by appropriate reasoning. Deep understanding of both principles of biology and the nature of scientific evidence are necessary in order to make informed judgments about the quality and appropriateness of evidence-based claims.
We developed and implemented an instructional model for introductory majors' biology that uses scientific models and arguments as both instruction and assessment. The course focuses on genetics, evolution, and ecology, and is the subject of a longitudinal study that examines the impacts of instructional reform on long-term student learning (NSF DUE 0736928). An overarching learning goal for the course was that students would build connections among biological concepts, rather than view the progression of content as a series of discrete and unrelated subjects. Our incorporation of concept modeling was an explicit strategy to force this way of thinking.
Early in the course, we provided instruction on concept modeling, including a guided discussion in which students derived the elements common to all scientific models – structures, behaviors and functions. Structures represent model components, behaviors indicate relationships between pairs of model structures, and functions describe the role or purpose of the model holistically (Goel and Chandrasekaran 1989; Hmelo-Silver et al 2000). In subsequent assessments – both formative and summative – we used SBF language to scaffold modeling activities and provide feedback about student work.
Throughout the course, we used a variety of cases and activities to provide students multiple opportunities to practice model construction and revise their models following feedback. As students constructed models, they wrestled with their understanding to determine which concepts were connected in meaningful ways and with how to explain the relevant processes that accounted for relationships. Early in the semester, students' models focused on building relationships among key genetics concepts (e.g., DNA, gene, allele, chromosome, phenotype). As the course progressed, students had to incorporate into their models increasingly complex principles of evolution (e.g., phenotypic variation, fitness, selection). It is important to note that students did not build a model that explained genetics concepts, then a new one to explain evolution. Instead, they progressively revised a "core" model that incorporated biological concepts as they learned them. Students adapted their models to explain how the same foundational principles applied to multiple cases, thus transferring "general" principles to "context-specific" cases.
It is a substantial cognitive challenge to build a conceptual model that accurately communicates how micro-scale molecular and sub-cellular genetic processes can lead to variation, expressed at the organismal level, which can be acted on by macro-scale processes in the environment to produce evolutionary changes in a population! It proved at least as challenging as asking students to construct a claim based on evidence and defend it using the principles in a scientific model (theirs or another). Indeed, student feedback at the midterm indicated a less than enthusiastic view for this approach to teaching and learning. However, by the end of the term, many students indicated recognition that this approach did, in fact, help them see that "every concept we learned in class [was] connected to the next concept after it" and "how all the topics were related - I could really see the connection between things that I thought were previously unrelated!"
Goel A, Chandrasekaran B. 1989. Functional Representation of Designs and Redesign Problem Solving. In Proceedings of the Eleventh Internation Conference on Articial Intelligence (pp. 1388-1394). Los Altos, CA: Morgan Kaufmann.
Hmelo CE, Holton DL and Kolodner JL. 2000. Designing to learn about complex systems. Journal of the Learning Sciences9: 247-298.