7.5 An Instructional Model

In concluding this chapter we will describe problem-based learning, an instructional model that we feel exemplifies the constructivist theory represented by our MAR metaphor. There are numerous instructional models popular today that focus on "problems." Case-based learning, modeled after the traditions of business and law school (Christensen, 1987; Spizzen & Hart, 1985; Stevens, 1983), is perhaps the most widespread and popular approach to problem-centered instruction (Greenwood & Parkay, 1989; Merseth & Lacey, 1993; Sykes & Bird, 1992; Wasserman, 1993). In this approach, instruction is centered around a description of some event that took place and that is relevant to the professional activities of the learners: an instructional scenario for teachers, a legal case for lawyers, etc. In business and industry, the instructional models for using problems are goal-based scenarios (Collins, 1994; Nowakowski, 1994) and action learning (Froiland, 1994). Both begin with a problem--a goal-or action-oriented decision the individual must make--but while the goal-based scenario model (see 20-33) uses problems from the past or specially created problems, action learning focuses on a real problem currently requiring action by one or more of the learners. It is, in essence, just-in-time training. Finally, at the elementary education level, the "problems" in the project-based learning model (Katz & Chard, 1989) involve a multidisciplinary (subject matter) exploration of a topic (e.g., trains) in which the students examine the topic from multiple perspectives over a week or more.

Problem-based learning as a specific instructional model was first implemented in medical education in the early 70s (see 20.3.4) and, like the models listed above, it too is based on presenting problems for the students to work on. It is not our goal to try to contrast the instructional models represented by cases/goals/projects/actions/problems. Indeed, there is so much variation in the implementation of each of these models that there is likely to be as much similarity between some implementations from different models than there is between different implementations within the same model (see, e.g., Barrows, 1986; Froiland, 1994; Williams, 1992).

A central theme to our chapter has been that an instructional designer's grounding assumptions about knowledge and learning are primary determinants of the instruction that is designed. So it is here. While there is widespread agreement as to the role of problems in instruction, the agreement is not so great in terms of the learning goals or assumptions about learning that surround the use of problems. We can identify five strategies for using problems that reflect different assumptions about either what is to be learned or how learning occurs.

I. The Problem as a Guide. Here the problem serves as a concrete reference point to focus the learner's attention. Reading assignments are given along with the case, and the readers are told to think of the reading in terms of the case. The case gives meaning to the reading assignment. This is similar to the study skill strategy of presenting questions at the start of the chapter to guide reading.

2. The Problem as an Integrator or Test. Here the problem is presented after the assigned readings are completed and perhaps even after they are discussed. The goal is to apply the knowledge from reading to the case to see how well the readings were understood and to aid the transfer process from learning to application. This is similar to presenting study or review questions at the end of the chapter.

3. The Problem as an Example. Here the problem is simply another bit of instructional material and is integrated in the reading. It is used to illustrate some particular point, and this is likely to be done through lecture or "teaching" as it is through student discussion. The focus is on the principle, concept, or procedure illustrated in the problem.

4. The Problem as a vehicle for Process. Here the focus on critical thinking in relation to the problem is central. The problem becomes a vehicle for training thinking skills. Thus, heuristics for problem analysis are taught in relation to the problem. The goal is to develop thinking skills, not only to solve the problem.

5. The Problem as a Stimulus for Authentic Activity. Here the focus is on developing the skills related to solving the problem as well as other problems like it. Rather than "teaching" the skills, the skills are developed through working on the problem, i.e., through authentic activity. The "skills" here include physical skills, gathering and bringing knowledge in the domain to bear on the problem, and metacognitve skills related to all aspects of the problem-solving process.

It is this last use of problems, the problem as a stimulus for authentic activity, that is our focus. It is the approach developed by Howard Barrows for medical education in the early 70s and which he has continued to develop and refine (Barrows, 1985, 1992, 1994). This approach is founded on the goal of engaging and supporting the learner in activities that reflect the demands of professional practice. Rather than "teaching" the student in the sense of presenting or even assigning information, the goal is to support the student's learning. From our perspective, the focus is rightly on the activity of the learner in the content domain. That is, it is impossible to describe what is learned in terms of the activity alone or in terms of the content alone. It is not that students are learning critical -thinking skills, self-directed learning skills, or collaborative learning skills, nor is it that they are learning "the" content domain. Rather, it is the activity in relation to the content that defines learning: the ability to think critically in that content domain, to collaborate with peers and use them to test ideas about issues, and the ability to locate information related to the issues and bring it to bear on the diagnosis.

We should emphasize that while problem-based learning (PBL) has been developed for professional training, it clearly has generality to all levels of education. A grounding assumption is that we do not learn in a content domain simply to acquire information but rather to bring that information to bear on our daily lives. Thus, consistent with the goals advocated by Dewey (1916, 1938), the argument in PBL is that learning in school should model and prepare us for the self-directed learning we will need to do to be effective participants in our community and to be effective in our profession. For example, in working with high schools, Barrows and Myers (1993) designed problems that related to the flooding in the Midwest in 1993, the action that government should take to monitor asteroids in space, and an analysis of how the geography of the Middle East impacts the conflicts between nations in that area.

The problems Barrows and Myers generated for the high school level may sound very much like topics that might be used in theme-based instruction. Indeed, PBL is theme based in the sense that learning is organized around the problem rather than around subject matter. However, once again we must emphasize that the critical characteristic for us is that the teacher in PBL does not teach students what they should do/know and when they should do/know it. Rather, the teacher is there to support the students in developing their critical-thinking skills, self-directed learning skills, and content knowledge in relation to the problem. The teacher must honor and support the students' thinking rather than impose structure on it. (We note, of course, that honoring the students' thinking will include challenging that thinking.)

7.5.1 The PBL Process

Problem-based learning can perhaps best be understood through a brief description of the learning/instructional process as implemented by Barrows (1985, 1992, 1994) in a medical school. When students enter the medical school, they are divided into groups of five, and each group is assigned a facilitator. The students are then presented a problem in the form of a patient entering with presenting symptoms. The students' task is to diagnose the patient and be able to provide a rationale for that diagnosis and recommended treatment.

The students begin the problem "cold"--they do not know what the problem will be until it is presented. They discuss the problem, generating hypotheses based on whatever experience or knowledge they have, identifying relevant facts in the case, and identifying learning issues. The learning issues are topics of any sort that are deemed of potential relevance to this problem and which the group feels they do not understand as well as they should. A session is not complete until each student has an opportunity to reflect verbally on his or her current beliefs about the diagnosis (i.e., commit to a temporary position) and assume responsibility for particular learning issues that were identified. Note that there are no prespecified objectives presented to the students. The students generate the learning issues (objectives) based on their analysis of the problem.

After the session, the students all engage in self-directed learning. There are no assigned texts. Rather the students are totally responsible for gathering the information from the available medical library and computer database resources. Additionally, particular faculty are designated to be available as consultants (as they would be for any physician in the real world). The students may go to the consultants seeking information.

After self-directed learning, the students meet again. They begin by evaluating resources: what was most useful and what was not so useful. They then begin working on the problem with this new level of understanding. Note that they do not simply tell each other what they learned. Rather, they use that learning in reexamining the problem. This cycle may repeat itself if new learning issues arise. Problems in the medical school program last anywhere from a week to 3 weeks.

Assessment at the end of the process is in terms of peer evaluation and self-evaluation. There are no tests in this medical school curriculum. The assessment includes self and peer evaluation (with suggestions for improvement) in three areas: self-directed learning, problem solving, and skills as a group member. While the students must pass the Medical Board exam after 2 years, this is outside of the curriculum structure.10 However, tests as part of the PBL curriculum are not precluded. For example, one high school teacher we know who uses the PBL approach designs traditional tests based, on what the students have identified as learning issues. Thus, rather than a prespecification of what is to be learned, the assessment focuses on the issues the learners have identified.

7.5.2 Key Issues in Designing PBL Instruction

7.5.2.1. Task Analysis. In designing a problem-based learning curriculum, as with any curriculum, we must begin with an analysis of what must be learned. However, in doing this, the developer must combine identification of the key concepts, procedures, etc., with an analysis of the professional (or "good citizen") use of those concepts. Identification of key concepts is a matter of expert statements of what is most important for students to "know." In both the Ohio University Business School (Miller & Stinson, 1993) and the Southern Illinois University Medical School (Barrows, 1985), this involved going to faculty teaching the traditional courses and asking them to identify the key things a student should learn in their course. This naturally requires extensive negotiation and specification. However, it does not involve the analysis of that key information into underlying learning requirements. That is, the task analysis stops at the top level and only identifies key concepts. What must be understood about the key concepts is defined through the professional activity that calls for their use; that is, it is defined in the activity of the learner. In the medical profession, the activity has to do with diagnosing and treating patients with presenting symptoms. In the business school, the professional activity has to do with business analysis and decision making.

There are two points we wish to emphasize here. First, this analysis does not preclude any type of learning activity--memorization of a list or extensive practice of a skill may be necessary--but it should arise out of the need to use the information in authentic tasks. Second, what must be learned includes not only information in the content domain but also metacognitive, collaborative, and other skills as are necessary for participating in authentic activity.

At Indiana University we have recently introduced an undergraduate minor in Corporate and Community Education (CCE) (Duffy, 1994). The goal of the minor is to develop the skills related to carrying out effective informal education related to community and professional needs. As part of the minor, the students take three core courses, all of which use a problem-based learning approach that involves the students in authentic educational problems. The skills and knowledge identified as critical in the CCE program and which guide the development of the problems in each of the core courses are outlined in Table 7-2.

TABLE 7-2. PRELIMINARY ANALYSIS OF THE SKILLS AND KNOWLEDGE REQUIRED IN THE CORPORATE AND COMMUNITY EDUCATION PROGRAM AT INDIANA UNIVERSITY I. Analyzing problems. Given a potential corporate or community education problem, how effective are you in analyzing that problem, deciding what needs to be done, and developing a plan of action- This includes your ability to: Work collaboratively in a group as both a leader and a group member, carrying your weight in the problem-solving activity, and listening to and respecting alternative points of view. Think critically about a problem, analyzing it into subproblems with some rationale. Evaluate alternative perspectives and prioritize the perspectives on the problem, including the perspectives of the various stakeholders. Design a work strategy addressing the sequence of activity, time requirements, and resource requirements. Use project-planning tools to manage your work. Monitor and adjust strategies as needed. II. Managing your learning. Given the analysis of a problem, how well can you identify and refine learning issues, locate resources relevant to those issues, and use those resources to obtain the information that will bear on the problem? This includes your ability to: Identify potentially relevant types of information resources and evaluate the usefulness of the resources after learning. Allocate the time necessary to achieve your self-directed learning objective. Soil through many relevant documents that express multiple perspectives, identifying relevant information and developing criteria for determining what information to use. III. Use of information resources. Given a learning issue, how efficiently can you use the variety of information repositories to identify and obtain potentially relevant information? This includes your ability to: Locate and acquire information or expertise from the library, experts, and using electronic resources like e-mail, World Wide Web, and Newsreaders. Reformulate your learning issue in a way appropriate to searching, using the particular information resource, i.e., ability to develop key words, restrict searches, identify related topics, etc. IV. Conduct audience/need analysis. Given a potential corporate or community education requirement, how well can you use the various strategies for evaluating the information needs? This includes your ability to: Determine what information needs to be collected about the audience. Design and evaluate alternative information collection strategies, e.g., phone interviewing, mail survey, door-to-door interview, questionnaires, etc. Implement an actual information collection strategy including determining strategies for sampling and accessing people. Analyze and summarize the results of the audience and needs analysis to make recommendations on information needs and delivery strategies. V. Designing and delivering usable information. Given the need to educate a group on some issue, how well can you use the necessary tools to design and deliver information that meets the information need? This includes your ability to: Develop a rationale for a delivery strategy, content specification, and content layout to meet the information need. Apply rhetorical, graphic design, document design, interface design, instructional design, speech communication, teaching, and adult literacy principles as appropriate to the preparation of the document. Use the various computer- and video-based tools as appropriate to the design and delivery VI. Assess effectiveness of your performance and your products. The goal here is to develop the ability to monitor and adjust as necessary your performance in each of the five areas described above. More generally, this goal can be phrased as one of becoming a reflective practitioner and includes your ability to: Reflect on your activities and evaluate strengths and weakness and, based on that, develop strategies to increase your effectiveness. Ability to solicit and use feedback from others on your performance. Ability to design and implement strategies for evaluating your products and for using that evaluation to assess alternative design or development approaches.

11 These are the skills and knowledge that the students should develop over the course of the program.

7.5.2.2. Problem Generation. The content for the course rests in the problem that is generated. It determines what the students must learn. There are two guiding forces in developing problems. First, the problems must raise the concepts and principles relevant to the content domain (as defined by the task analysis). Second, the problems must be 11 real." There are three reasons for this. (1) Because the students are open to explore all dimensions of the problem there is considerable difficulty in creating a rich problem with a consistent set of information. (2) Real problems tend to engage learners more; there is a larger context of familiarity with the problem. (3) Students want to know the outcome or current status of the problem and tend to be disenfranchised when told it is not a real problem.

In the case of the medical school, a real problem means the case is based on a real patient, not necessarily a current patient, but not someone fictitious for whom symptoms are made up. In the business curriculum, this means that the problem is a current business problem; e.g., 3 years ago a problem that was meant to engage the students in particular concepts was "Should AT&T buy NCR?" The parallel problem last year was "Should Merck buy Medico?" (Stinson, 1994). For our corporate and community education course, the first problem, one that will consume an entire semester, is stated as follows:

About 15 years ago in Bloomington, a sludge byproduct of Westinghouse's manufacturing process was distributed around the community as a mulch rich in nitrogen. However, it was soon determined that the sludge was contaminated with PCBs, a chemical thought to be a significant carcinogen. The sludge has been gathered into piles and covered with tarps or concrete, though there are thought to be numerous sites around the county that are still contaminated.

For the last 15 years there have been numerous proposals as to how to dispose of the PCBs, but no action as yet has been taken. Interestingly, the public has been apathetic about this potentially very serious issue. This past summer the EPA set up eight, 2-hour public meetings to discuss disposal of the PCBs. However, because of lack of response, they cut it back to two meetings, and then only 14 people showed up at one of those meetings. Another forum held this November was also sparsely attended. In both cases, the meetings received minimal press coverage.

As responsible members of this community we are concerned that the PCBs are still scattered around Bloomington after all these years, and the public does not seem to care. We are going to design educational materials that will:

• Provide the information the citizens need to make a reasonable decision or to discuss alternatives that will lead to cleaning up the PCBs
• Present that information in such a way that the citizens will be motivated to actively participate in decision making regarding the PCBs.

7.5.2.3. The Learning Sequence. In this sequence, problem-based learning cycles go through two types of learning activities: collaborative problem analysis sessions and self-directed learning. The collaborative problem analysis session usually occurs with small groups of about five, supported by a facilitator. However, with modifications, this can be a mixture of large- and small-group activities (Miller & Stinson, 1993). In our corporate and community education program, we will use a jigsaw model. In the first stage of the jigsaw, the focus is on developing content expertise related to the problem area--PCBs in the case of the problem outlined above. The whole class will first work to identify learning issues in the content area, and then small groups assume responsibility for particular issues. They develop expertise on those issues and then share that expertise in large-group problem solving where all the expertise is brought to bear on the problem. Then as the class moves to the second stage of identifying the instructional strategy and designing the product, the groups will be redefined so that product development groups will consist of students having content expertise on different issues.

The sessions are student-run, problem-solving sessions in which hypotheses and action plans are generated, along with facts that support or refute the hypotheses and learning issues that must be addressed after the session. The learning goals underlying the design of these sessions include developing informal or hypothetico-deductive reasoning skills, reflective and metacognitive skills, and collaborative skills and content knowledge as other members of the group bring their content knowledge to bear on the problem.

In self-directed learning, the students seek and use resources that will address the issues they need to learn about. This is considerably different from learning activities in traditional instruction, where the reading is assigned by the instructor and the task is to learn what is in the text. In the PBL format, the students are learning how to identify, locate, and evaluate information resources as well as use those resources as tools in solving problems, rather than as ends in and of themselves. Interestingly, at all levels of schooling, the problem-based learning tends to lead the learners to primary sources, with secondary sources like textbooks being rejected as too vague or not current (Barrows, personal communications; Duffy & McMahon, 1992).

7.5.2.4. Facilitator Role. In his discussion of the tutorial process, Barrows states:

The ability of the tutor to use facilitory teaching skills during the small-group learning process is the major determinant of the quality and success of any educational method aimed at (1) developing students' thinking or reasoning skills (problem solving, metacognition, critical thinking) as they learn, and (2) helping them to become independent and self-directed (learning to learn, learning management). Tutoring is a teaching skill central to problem-based, self-directed learning" (1992, p. 12).

Throughout a session the facilitator models higher-order thinking by asking questions that probe students knowledge deeply. To do this, the facilitator constantly asks Why? What do you mean? How do you know that's true? The facilitator's interactions with the students remain at a metacognitive level, and he or she avoids expressing an opinion or giving information to the students.

A second facilitator role is to challenge the learner's thinking. The facilitator (and hopefully the other students in this collaborative environment) will constantly ask: Do you know what that means? What are the implications of that? Is there anything else? Superficial thinking and vague notions do not go unchallenged. During the first few PBL sessions, the facilitator challenges both the level of understanding and the relevance and completeness of the issues studied. Gradually, however, the students take over this role themselves as they become self-directed learners.

7.5.2.5. Assessment. Assessment must be in the context of the problem the students are working on. There are numerous strategies for accomplishing this. Mildred Jackson, a high school science teacher at Choctaw County High School in Butler, Alabama, uses problem-based learning for the majority of her science instruction--with the reading materials being located by the students based on their leaming issues. That is, there is no assigned reading. Butler simply uses for her testing what the students identify as learning issues. The text may be multiple choice, essay, or short answer; the critical characteristic is that it is generated from the students' learning issues.

Barrows relies entirely on student self? and peer assessment. However, the assessment is ongoing, rather than just being an end of the semester rating. Thus, while the students may be easy on one another initially, as they continue working on problems, they clearly begin to provide ratings and feedback to fellow group members more reflective of their contributions.

We also rely heavily on peer and self-assessment in the corporate and community education program. As noted in Table 7-2, the abilities to self-assess and to provide constructive feedback to team members are explicit learning goals, and this is not only an assessment process but also a learning process. Every other week, students assess themselves in terms of their skills as a problem solver, as a self-directed learner, and as a team member. They are provided guidance on the skills and knowledge development issues (Table 7-2) they should be considering in each of these categories. They also evaluate themselves by describing their learning over the last two weeks.

The self-evaluations are distributed to the instructor and to the team members. During the following week (the alternating weeks), the team members will provide feedback to the other team members as to how they performed--that is, they will examine the self-assessments for each team member and offer feedback, including suggestions on how they might improve their performance. These peer evaluations are sent to the instructor, who integrates them and presents them to the individual students, thus providing anonymity in the peer evaluation process. While we would prefer open evaluations, we recognize that it is something we must work toward over the course of the curriculum.

The Indiana University Northwest Medical Center presents an interesting variation on the assessment strategy. While it has a PBL program, because it is part of a larger system, it was required to administer unit tests every 6 weeks that "covered" the subject-matter focus for that period. In discussions the first author had with the instructors, they reported that students worked hard and enjoyed the problems, but that after each problem some of the students would pester the instructor, asking, "That was all very good, but what do I really have to know for the test?" This became an ongoing problem for the instructors. In response, the dean of the center took the test development out of the hands of the instructors. Rather than developing a test, they purchased the unit test from another medical school, a different test each time so that even the instructors did not know what school the tests came from.13