The Question
Does a project- or design-based approach to teaching about complex systems result in greater student learning than traditional instructional methods?
The Context
Problem-based learning, project-based learning, and Learning by Design are all variants of a constructivist learning model that emphasizes guided student engagement in discovery learning. Each variant seeks to present students with a complex problem or challenge, requiring student identification of and engagement in the knowledge and skills necessary to successfully solve the problem or address the challenge. Essentially, students use the scientific process of discovery (observation, theory building, experimentation, data analysis, drawing conclusions) to solve problems, theoretically internalizing knowledge to a greater degree because they have actively engaged in the learning process, rather than simply receiving the solutions and context in traditional lecture or presentations. Critics of such discovery learning worry that students will not engage sufficiently in the content of the curricular domain, and even supporters have acknowledged that such in-depth learning may result in uneven engagement with the current broadly defined curriculum. Although problem-based learning has been used extensively in medical schools and higher education, it has received less research attention at the K–12 level.
The Details
Cindy Hmelo, Douglas Holton, and Janet Kolodner conducted the study highlighted in this issue of ResearchBrief (see below for full citation). The researchers wanted to see if a design-based approach to a unit on the human respiratory system would result in greater student learning than more traditional direct instruction methods. The researchers had three specific goals:
- Explore the extent to which a design-based approach could help students learn about the human respiratory system.
- Examine whether the structure-behavior-function (SBF) framework of design-based learning would help students think deeply about the topic.
- Investigate implementation of Learning by Design methods in a middle school setting.
Although goals one and three are relatively straightforward, the researchers took some time to explain the concept of “deep” learning mentioned in goal two. Because the human respiratory system is so complex, the researchers noted that learning could occur on multiple levels; for example, students might learn about the components of the system without really understanding the interrelations of the pieces or the overall functioning of the system. They noted that phenomena in biology can occur at the anatomical, biochemical, and physiological levels simultaneously, making deep understanding difficult to achieve. By using the SBF approach, the researchers hypothesized that learning could occur at the following levels:
- Structural: the physical structure of the lungs within the system; the structure of the alveoli within the lungs
- Behavioral: the movement of the chest or of air going in and out of the nose; more complex behaviors, such as the passage of gas from high to low concentration across the semipermeable membrane of the lungs
- Functional: lungs move to bring in (or expel) air; oxygen is brought in for cells; waste is removed from cells
By creating a learning experience where students actually design an artificial lung, the researchers hypothesized that students would gain a deeper and more functional understanding of the respiratory system than would students taught with a more traditional lecture and direct instruction method. One of the challenges the researchers faced was designing a project that engaged students but also included time for substantive reflection that would support a deep understanding of the data collected through the experience. As a result, they combined parts of problem-based learning with case-based reasoning to create what they call Learning by Design (LBD), a complex teaching model that allows students to engage in true problem solving. An important aspect of LBD is its use of failure (at a task) as a learning opportunity and motivational tool: students create something, test it, gather data and get feedback, explain outcomes and anomalies, and revise expectations and try again.
The researchers studied 42 students in two 6th grade life sciences classrooms meeting over 13 consecutive class periods. The students in these two classes were taught about the human respiratory system using the LBD model, while a third control classroom was taught using traditional methods (textbook readings, lectures, and teacher-directed discussions.). The researchers conducted 20 pre-intervention interviews and gave all students a 12-item true-false test to gauge initial knowledge of the respiratory system. The classes were videotaped and observed daily, with the researchers maintaining detailed research logs. The teacher in the treatment classes was described as inventive, energetic, knowledgeable, curious, and reflective, although he had never taught using LBD before.
In analyzing the classroom experiences of the teacher and students, the researchers identified a variety of mixed outcomes. Student reflection was generally well managed throughout the learning process, and the LBD teacher successfully used scaffolding to support learning; however, the transition to helping students design and debug their models was minimal. Although student motivation was high and students were engaged, the complexity of the task (and the restricted timeframe) meant that students had limited opportunity to apply what they had learned. The researchers also noted a disconnect between students' initial research and their project designs and observed strong group-level collaboration within the groups, within the classroom, and across groups. These student interactions supported progress throughout the classroom.
In addition to these descriptive outcomes, the researchers gathered data related to specific learning outcomes. They administered 12-item pre- and post-tests measuring student conceptual understanding of the respiratory system, evaluated student illustrations of the respiratory system within the outlines of an empty human body, and conducted clinical interviews. The control class was also assessed. The LBD class showed significant improvement between their pre- and post-tests, while the control class did not. Based on analysis of the students' illustrations, the LBD students had significantly more advanced conceptual understanding of the respiratory system.
For teachers interested in implementing LBD in their classrooms, the researchers suggested that teachers take the time to introduce the process over multiple lessons to help students become comfortable with the activities. They also made seven other recommendations:
- Present the challenge with appropriate and complete materials.
- Engage in a prelearning activity involving related artifacts that can allow students to explore issues they are likely to face when engaged in the modeling part of the LBD activity.
- Use regular classroomwide reflection and planning to help students broadly explore ideas, issues, and plans.
- Allow groups to address component issues of the problem and present their findings to the class to help build a holistic understanding of the problem and design solutions.
- Have groups regularly reflect on new knowledge and revise their design plans (sharing their thinking with the class).
- Have students reflect and revise designs based on testing of the created models, again sharing results with the class.
- Have students work toward final presentation, reflection, and review.
The Bottom Line
Students engaged in a hands-on Learning by Design educational experience were highly motivated by the challenge and—through extensive research, group work, and classroom discussion—reached a deeper and more complex understanding of the respiratory system than did their peers receiving traditional instruction in a comparison classroom.
Who's Affected?
The focus of this research was students in a middle school science classroom.
Caveats
The researchers were careful to highlight the fact that Learning by Design is not a panacea and is extremely difficult to implement well. The process is complex and requires skilled facilitation by the teacher to ensure the learning outcomes are adequately addressed. The researchers suggest that they should have included a prelearning activity (such as unstructured time for students to engage with spirometers or pumps) to help students bridge the gap between researching a problem and designing a solution. They also noted that the materials they provided actually limited student models because they didn't supply any filtering material (needed to represent gas exchange). Although the pre- and post-tests favored the intervention group, the pre-intervention interviews of the control group were lost due to a “technological mishap,” making it impossible to compare the improvement for students in the traditional classroom on some measures. Additionally, the instruction that the control group received was not well described, making it difficult to evaluate the comparative worth of the sample. Both the control and intervention groups were small, and, as such, the results may not be generalizable to other students, classrooms, and teachers.
The Study
Hmelo, C. E., Holton, D. L., & Kolodner, J. L. (2000). Designing to learn about complex systems. Journal of the Learning Sciences, 9(3), 247–298.
Other Resources
Learning by Design
Georgia Institute of Technology
Problem-Based Learning Network
Illinois Mathematics and Science Academy
Project Based Learning Handbook
Buck Institute for Education
