Virtual Design Center

PRINCIPLE 12:
Providing Scaffolding for Complex Cognitive Skills Helps Learners Engage in Expert-like Thinking Process

The original paper, entitled “Case Application Suite: Scaffolding Use of Expert Cases in Middle-school Project-based Inquiry Classrooms,” was presented at the 2004 International Conference of the Learning Sciences in Los Angeles, California.[1]

The paper is useful for science, technology, engineering, and mathematics (STEM) instructional designers because it illustrates the value of several core ideas among scientists who study learning and cognition. The study was one of a series conducted in the context of a project-based middle school science curriculum known as Learning by Design (LBD). In contrast to the many project-based science curricula that are organized around more conventional scientific problems, LBD units are structured around challenges more akin to the kinds of problems solved by engineers. Within the widely known framework for teaching and learning known as cognitive apprenticeship, the study used the theoretical foundation of case-based reasoning to refine project-based science curricula. In a unit called Tunneling Through Georgia, the study demonstrated the value of an approach to scaffolding that significantly improved student learning of core science topics in geography. This scaffolding was provided by a software tool known as Case Application Suite (CAS) that supported “just-in-time” provision of modeling and coaching while students worked on the project-based curricular unit.

This study builds on Kolodner’s (2002) prior research on case-based reasoning. The power of case-based reasoning is perhaps most obvious in domains such as law and medicine, where most of the knowledge that defines the disciplines are actually represented as cases. Among STEM professionals, the continual advances in engineering knowledge are a good illustration of the power of case-based reasoning and learning.

Case-based theory nicely complements the widely known cognitive apprenticeship perspective on teaching and learning (Collins, Brown, & Newman, 1989). The scaffolding of learner activity at the heart of cognitive apprenticeships lets learners see to-be-learned skills modeled in a relevant context, and then be coached to use the skills, and then use the skills repeatedly in varied contexts. The coaching involves prompting, hinting, and providing reminders as learners are carrying out targeted skills to solve problems. In addition to learning how to carry out skills, learners must know when skills are needed and the right sequence for using them. To do so, students need multiple opportunities to engage in tasks and be provided feedback that they then use to refine their performance. Support for the cognitive apprenticeship approach is provided by extensive research on the transfer of learning—the use of skills and knowledge developed in one context (the learning context) to some other context (the transfer context). Studies of transfer show that flexible (i.e., useful) knowledge is developed when learners have opportunities to carry out skills in a variety of situations where those skills are appropriate. Studies also show that carefully prompting students to use skills in learning environments makes it easier to learn the skills that are necessary to succeed in a range of transfer environments.

Kolodner and others have been exploring the power of case-based theory and project-based methods for creating cognitive apprenticeships for K-12 science education. These efforts engage students in extended projects around cases that embody targeted established science education standards. Kolodner’s research involves a middle school science curriculum known as Learning by Design (LBD; Kolodner et al., 1998, 2003). LBD has students learn science content and scientific reasoning in the context of “design challenges.” A distinct aspect of LBD is that these challenges are more akin to the problems solved by disciplinary engineers. In each LBD unit the teacher first models the design of experiments, the interpretation of data, using data as evidence and explaining scientifically. Students are then coached to apply those skills in small collaborative groups working together to solve a variety of project challenges. Students are provided access to expert cases to provide guidance in solving the challenges. Each group’s solution is then examined and discussed by the class in a variety of present-and-share venues.

This particular study tested the value of an innovative software tool called the Case Application Suite. The tool was designed to help teachers model expert reasoning and coach students to reason like experts. It does so by sharing responsibilities with the teacher as students are learning to interpret and apply expert cases, providing modeling, coaching, and scaffolding for learners while working in small groups. The key element of this tool is that it supports just-in-time scaffolding. The study predicted that learners in a cognitive apprenticeship who have just-in-time scaffolding available as they are carrying out skills in small-group settings will learn those skills and associated science content more competently than those in the same cognitive apprenticeship environment who do not have that scaffolding available.

This study addresses two challenges in collaborative project-based inquiry. One is that small groups need help in being intellectually rigorous and productive. Without the help of the teacher as they are working, students sometimes struggle to focus their attention appropriately on the reasoning they need to do. To address this issue, Kolodner’s team had previously designed a suite of software tools called SMILE (Kolodner & Nagel, 1999). SMILE’s scaffolding fills in for the teacher as students are working in small groups, providing organizational help, hints, and examples for each reasoning skill being used. Second, for many science topics (particularly Earth science topics like geology and tunneling), it is impossible for classroom learners to design and build the working artifacts (e.g., tunnel walls and air shafts) that are central to project-based inquiry. This challenge is addressed by providing alternative artifacts around which learners can identify issues they need to learn more about. For example, to understand how core sampling can allow geologists know what kinds of rock exist in different parts of a mountain, they take “core samples” using a straw and different colors of stacked Play-Doh™. The Tunneling Through Georgia unit promotes identifying issues by having learners read and use expert cases. To help them understand and apply these expert cases, they created a suite of tools within SMILE called the Case Application Suite (CAS, Owensby & Kolodner, 2002).

In the professional development associated with LBD, implementing teachers are provided with teacher materials to help them model the reasoning involved in interpreting and applying cases. During the unit teachers model interpretation reasoning for students and help them identify the ins and outs of their own reasoning about cases while sharing interpretations with their peers. When the teachers implement each LBD unit, the CAS scaffolds the students’ enactment of these reasoning strategies. Using examples, prompts, hints, and organizational charts, CAS plays the role of coach as individuals or small groups of students are working to interpret an expert case and apply it to their challenge. Its questions prompt students to identify the lessons they can glean from a case, analyze their applicability to the new situation, consider alternatives and next steps, and assess the goodness of their proposals. Hints and examples are provided with each prompt to give students clues about the kinds of things they should consider and articulate in their reasoning.

CAS includes three tools. The Case Interpretation Tool (Figure 1) helps students identify problems the experts encountered in achieving their goals, solutions they attempted and why they chose those, what criteria and constraints informed those solutions, what results they accomplished and explanations of those, and any design rules of thumb (lessons learned) that can be extracted from the experience. The Case Application Tool guides the application of the design rules of thumb gleaned from the expert case. Students are prompted to consider, for each design rule of thumb, if it is applicable to their challenge and the different ways it could be integrated into their solution. The Solution Assessment Tool helps students make predictions about the success of their proposed solutions, analyzing the impacts they expect their solution to make as well as where they expect their solution to fall short. As shown in Figure 1, each tool has a left frame that displays the case students are analyzing or summaries they’ve written of that case; a middle frame with prompts (questions); and a right frame that displays hints, examples, comments, and templates.

CAS’ tools provide scaffolding in a variety of ways. First, the three tools that make up the suite provide a high-level scaffold because each tool corresponds to a major step in the case application process. Second, the prompts in each tool's center frame focus students’ attention on important aspects of the case (e.g., the problems the experts faced, the solutions they created, the criteria and constraints that informed solutions, and rules of thumb that can be gleaned). Third, with each of those prompts, hints and examples are provided to give more specific help. Finally, charts and templates serve as organizers to help students create and analyze the applicability of rules of thumb they have extracted.

In addition to providing scaffolding to complement coaching provided by the teacher, CAS provides a second forum for students to share their interpretations and applications with each other. When integrated well into a classroom, the teacher models the stages of case use and coaches students as a class as they attempt to interpret or apply a case or assess a solution. Small groups then use the software as they carry out those processes without the teacher. Then groups present their cases to the class followed by a teacher-facilitated discussion about what can be learned from the full range of cases and the reasoning the most successful students engaged in to do their work. Then students edit their case reports based on discussions and post them for use by the entire class.

This study examined the Case Application Suite within an implementation of the Tunneling Through Georgia unit implemented in two classes by a sixth grade teacher. In Tunneling Through Georgia groups of three or four students help design a transportation tunnel across the state of Georgia. Four tunnels need to be designed, each for a different geological area of the state—mountainous, a sandy region, and so on. They need to address several issues—at what depth to dig the tunnel, what methods to use for the digging, and what support systems are needed in the tunnel's infrastructure. Cases suggest which geological characteristics of the tunnel location they need to learn more about to address the challenge, introduce students to different kinds of tunneling technologies, and give them an appreciation of the complexity of tunnel design. Cases also provide examples of approaches that worked well that they can apply to their designs and approaches that did not work well so they can avoid making the same mistakes. Planning to avoid those mistakes usually requires learning more about the geology of the tunnel region. Data was analyzed from two of Mrs. K’s classes. In each class some student groups used the software (n=14 students; 4 groups), and the remainder used only the structural kind of scaffolding provided by the My Case Summary design diary page (n=33 students; 9 groups). My Case Summary design diary pages have a chart with four columns: Case Summary, Problems that Arose, How Problems Were Managed, Ideas for Applying to Our Challenge.

The researchers compared the activity and capabilities of the two groups of students and conducted interviews in order to answer three research questions: (1) How are students’ abilities to interpret and apply cases to their project challenge affected by such scaffolding (effects with)? (2) To what extent would students’ ability to apply cases in the absence of the tool be influenced by its use during a project (effects of)? (3) To what extent does the tool enable students to articulate the processes involved in case application? They predicted that that the combination of teacher modeling and software coaching should result in students’ interpreting and applying expert cases to their challenge in more sophisticated ways. In particular, this prompting and hinting should result in students creating more useful and descriptive interpretations and applications of expert cases to their challenge. Because of the help they get while reasoning about cases, students exposed to CAS’ scaffolding should also be more cognizant of their reasoning, making them better able to reason with cases when the tool is not available and better able to articulate the steps involved in reasoning about cases. In particular, given the tasks that CAS scaffolds, we predicted that students who had use of the software during small group work would be better at things like making connections between problems and solutions in the expert case, understanding and identifying criteria and constraints in the expert case, making claims or arguments using the expert case as justification, etc.

The implementation of Tunneling Through Georgia. Mrs. K’s enactment of case interpretation and application was very much a cognitive apprenticeship approach. Earlier in the year, Mrs. K modeled case application skills for students as they learned about erosion, and she coached them in small groups as they applied erosion management methods to achieve the challenge of keeping a hill from eroding on a basketball court. At the beginning of Tunneling Through Georgia, Mrs. K coached the class through the reasoning involved in analyzing another case design diary page that she had created consisting of columns labeled Facts, Problems, Solutions, Constraints, Ideas/Rules of Thumb, and Questions. At the end of this exercise, she informed the class that they would do the same thing with the cases that they would be assigned, and she assigned each group to work on a case for three class periods (with some of the students using the CAS and the others working without it). Each group made a poster to present their case to the class during six class periods devoted to case interpretation and discussion. Ten days later three class periods were devoted to applying the expert cases to the solutions they were designing. The CAS groups used the tools to analyze the rules of thumb they gleaned from cases they read, brainstorming all the ways that each rule of thumb could inform their solution and determining which application made the most sense given their group’s goals, criteria, and constraints. The non-CAS students used the teacher’s paper template to analyze their challenge and design a solution, while the teacher assisted groups as needed. Each group then wrote reports giving their recommendations and justifications for their sections of the tunnels.

The researchers videotaped the students presenting their cases and systematically analyzed the scientific quality of the presentations along with the scientific quality of the posters each group produced. After the implementation each group also completed a performance assessment where they investigated the feasibility of building two subdivisions on an island off the coast of Georgia. The task was similar to the prior challenges, but the only scaffolding that was provided consisted of a chart that students had to fill in. While they discussed the expert case as a team, each student was responsible for completing a chart individually. Group discussions were videotaped and analyzed along with individual written responses for evidence of student reasoning.

How Did This Learning Technology Help Learners Internalize the Complex Cognitive Skill?

The quality of the presentations, and posters in Mrs. K’s class was very high among both the CAS and non-CAS groups. But careful analysis showed additional qualities among the CAS group that were strongly related to the goals of the CAS tools. First, the CAS students more clearly recognized that addressing constraints on their design would lead to positive outcomes and that ignoring constraints would lead to negative outcomes. Second, the CAS groups demonstrated more sophisticated models of causality in their rules of thumb (“Take core samples—they can save your life because if you hit the wrong kind of rock”) compared to the non-CAS group (“Take core samples”). Third, CAS groups tended to include the process the experts used to implement the solution, whereas the non-CAS groups generally left out these important aspects of expert reasoning. Analysis of the video recording of the groups collaboratively completing the performance assessment also showed significantly higher quality (mean rating differences p < .05) collaborative reasoning for the groups who used the Case Application Suite. Specifically during the first part of the task, CAS groups were more likely to (1) describe expert problems on a finer grained level, (2) discuss whether a management method made sense, and (3) use the cases to understand the context in which problems arose. In the second part of the task, the CAS groups (1) tended to identify issues or problems not explicitly stated in the case that were important to their group’s challenge, (2) to carry over the relevant aspects of the case that they identified in the first part, and (3) to justify the use, modification, or abandonment of an expert solution based on the criteria and constraints of the group’s challenge.

Analysis of each group’s written responses also revealed advantages for the CAS groups, but none of these reached statistical significance. The CAS groups (1) identified more expert problems that were the result of experts implementing a solution, (2) better identified the pros and cons for each management method they identified, and (3) provided more correct justifications of risks.

Interviews with the 14 students in the CAS groups indicated that the software (1) helped them organize their thoughts, (2) identify important aspects of the cases quickly, (3) keep track of where they were in the bigger task, (4) pull out important facts, (5) formulate more thoughtful answers, and (6) write facts and answers on their own. All but one of the students was able to clearly articulate the various functions of the software and the embedded relationship between the software and the cases.

At a very specific level, this study showed that the just-in-time scaffolding in the Case Application Suite helped students interpret and apply expert cases. Having the software available when working in small groups without the teacher seemed to help students engage in more focused and detailed reasoning. If teacher and software share their approaches to carrying out the targeted skills, then software can be used fairly naturally as an agent with expertise working with small groups, allowing the teacher to successfully share scaffolding responsibilities as students learn complex cognitive skills. This study also suggested that the kind of help provided by CAS promotes internalizing the processes involved in carrying out complex skills. Learners generally had a good idea of the steps involved in the overall process of applying cases to new situations. We need to find out which parts of the scaffolding system are responsible for this.

At a more general level, this study illustrated the value of case-based theory and project-based methods for creating cognitive apprenticeships for middle school science students. Both the CAS and non-CAS students demonstrated sustained engagement in the sorts of scientific inquiry that science teachers strive for but seldom achieve. The study also illustrated the potential of using engineering design challenges for project-based learning in science disciplines (rather than more conventional problems that directly target disciplinary science problems). The increased outcomes for the CAS groups illustrate the value of just-in-time scaffolding. The study provides both theoretical framework and inspiring examples for providing useful scaffolding in a wide range of STEM instructional contexts.

[1] This paper was a winner of the 2005 Virtual Design Center's paper competition, selected through an extensive review process by the Virtual Design Center advisory board. The above summary was prepared by the authors with the assistance of the advisory board chair (Daniel T. Hickey) and the project manager (Beaumie Kim). For more information on how you could use this design principle for your practice, you may contact the Virtual Design Center (vdc@cet.edu) or the board representatives (Daniel T. Hickey, dthickey@indiana.edu; Beaumie Kim, bkim@cet.edu) and the authors (Jakita N. Owensby, owensby@us.ibm.com; Janet L. Kolodner, jlk@cc.gatech.edu).