23.4.1.1. Historical Antecedents. REALs (i.e., constructivist learning environments, information-rich learning environments, or interactive learning environments) are not new to education. We can go back to Socrates (470-390 BC) and see that he used problems and questions to guide students to analyze and think about their environments (Coltrane, 1993). Rousseau prescribed using direct experience (Famham-Diggory, 1992). In the early 1900s, John Dewey (1910) proposed student-directed reforms and experiential learning. Bruner (1961) advocated discovery or inquiry learning around realistic problems. The notion that students should learn through practice, application, and apprenticeship has been with us for centuries. It wasn't until the industrial age, when we needed places to store children until old enough to work on assembly lines, that we began trying to mass produce replicable results. Yet, the last 5 to 10 years have seen renewed emphasis on reforming schools and teaching practices to replace our production lines with classrooms that teach people to think and to solve problems. This current effort at renewal revolves around a set of ideas and theories referred to as constructivism.

23.4.1.2. Characteristics of Constructivism. The class of theories that guides the development of REALs is called constructivist theories (see 7.3; Bednar, Cunningham, Duffy & Perry, 1991; Bransford & Vye, 1989; Clement, 1982; Duffy & Bednar, 1991; Minstrell, 1989; Perkins, 1991; Resnick & Klopfer, 1989; Scardamalia & Bereiter, 1991; Schoenfeld, 1989; Spiro, Feltovich, Jacobson & Coulson, 199 1). 1 shall, for background, provide only a brief description. Fundamentally, constructivism asserts that we learn through a continual process of constructing, interpreting, and modifying our own representations of reality based on our experiences with reality (Jonassen, 1994c). Learning includes a social component and conceptual growth comes from sharing perspectives and modifying our internal representations in response to that sharing (Bednar et al., 1991). Wheatley (1992) summarizes these ideas and emphasizes the importance of active involvement in the environment:

Learning (innovation) is fostered by information gathered from new connections; from insights gained by journeys into other disciplines or places; from active, collegial networks and fluid, open boundaries. Learning (innovation) arises from ongoing circles of exchange where information is not just accumulated or stored but created. Knowledge is generated anew from connections that weren't there before. When this information self-organizes, learning (innovation) occurs, the progeny of information-rich, ambiguous environments (p. 113).

There are several important characteristics of the constructivist view of learning that govern our design of REALs. First is the notion that knowledge is not a product to be accumulated but an active process in which the learner attempts to make sense out of the world (Gurney, 1989) that is under constant evolution. Brown, Collins, and Duguid (1989) illustrate this idea:

A concept, for example, will continually evolve with each new occasion of use, because new situations, negotiations, and activities inevitably recast it in a new, more densely textured form. So a concept, like the meaning of a word, is always under construction (Brown et al., 1989, p. 33).

A second characteristic is the notion that people conditionalize their knowledge in personal ways (Gurney, 1989). That is, they acquire knowledge in forms that enable them to use that knowledge later (Bransford et al., 1990). Bransford (1990, p. 122) states that "... there are large differences between knowing something and spontaneously thinking to do it or use it when one is engaged in an actual problem-solving situation." Knowledge is "indexed" to the contexts in which we encounter it. We are unlikely to use knowledge that is decontextualized because it has no relevance for us. A person who indexes and conditionalizes knowledge knows when to apply that knowledge. A person who learns in a decontextualized way often is not aware that he or she has the applicable knowledge to solve a problem. Students must acquire concepts and theories in ways that help them use the information later on and appreciate the value of that information. Brown, Collins, and Duguid (1989, p. 36) describe this process as "indexicalizing knowledge." They mean that rich involvement in* realistic and relevant problem solving enables learners to develop many broad and deep indexicalized representations that enable them to apply more spontaneously knowledge to new situations because they can compare a known and relevant situation with a new situation. The more links there are across related knowledge structures, the more likely students are to apply that knowledge. The more closely the learning context resembles the actual context, the better people will perform. Tulving and Thompson (1973) refer to this as encoding specificity, which holds that successful retrieval of information is enhanced when cues relevant to later retrieval of that information are encoded along with the material learned. It is important to note that constructivists contend that these rich links cannot be developed in decontextualized learning activities; rather teaming must be placed in realistic contexts that provide cognitive conflict or puzzlement and determine the organization and nature of what is learned (Savery & Duffy, 1994).

White (cf. Robertson, 1990) theorizes that there are two kinds of links that need to be developed while learning: internal and external. Internal associations are connections among the criterial attributes of a principle. Internal associations reflect the learner's understanding of the concept. External associations refer to connections between the principle and everyday experiences or context and indicate the "usability" of a concept. Learning to solve problems requires both kinds of links. Our schools are good at building the internal links, but poor at providing the external links.

The third major characteristic of constructivism is the importance of collaboration and the social negotiation of meaning. Common understandings and shared meanings are developed through interaction among peers and teachers. This is the cultural aspect of knowledge.

The activities of a domain are framed by its culture. Their meaning and purpose are socially constructed through negotiations among present and past members. Activities thus cohere in a way that is, in theory, if not always in practice, accessible to members who move within the social framework. These coherent, meaningful, and purposeful activities are authentic, according to the definition of the term we use here (Brown et al., 1989, p. 34).

This social aspect of constructivism is important on an individual level as well as cultural level, for collaborative interactions allow us to test the viability of our understandings, theories, and conjectures (Savery & Duffy, 1994).

23.4.1.3. REAL Example: Cognitive Flexibility Theory. Cognitive flexibility theory (CFT) implements many of the ideas of constructivism, (Jacobson & Spiro, 1992; Spiro et al., 1991), particularly focusing on the development of conditionalized and indexicalized knowledge structures.

Essentially, the theory states that cognitive flexibility is needed in order to construct an ensemble of conceptual and case representations necessary to understand a particular problem-solving situation. The idea is that we cannot be said to have a full understanding of a domain unless we have the opportunity to see different case representations (Borsook & Higginbotham-Wheat, 1992, p. 63).

CFT attempts to teach content in ill-structured domains, that is, in domains where the knowledge base is so vast and complex that multiple solutions to problems are possible and likely. There are no clear-cut answers in ill-structured domains, so simple algorithms often fail. 111-structured domains include law, medicine, and education. Therefore, CFT emphasizes the following instructional strategies to help learners develop rich and deep knowledge structures (Jacobson, 1994):

1. CFT uses several cases and rich examples in their full complexity. One of the tenets of CFT is to avoid oversimplifying knowledge and examples because it leads to future misunderstandings that are difficult to change.
2. CFT uses multiple forms of knowledge representation, providing examples in several kinds of media. CFT encourages students to look at knowledge in several ways and from several perspectives.
3. CFT links abstract concepts to case examples and brings out the generalizable concepts and strategies applicable to other problems or cases.
4. To avoid the mistakes of oversimplification, CFIF presents a number of examples to make apparent, rather than hide, the variability of concepts and themes within the domain.

Although CFT does not emphasize the collaborative nature of knowledge construction, its strategies include opportunities for social negotiation of meaning. We examine further on in the chapter other examples of REALs that place a high emphasis on cooperative learning.

The second characteristic of a REAL is that learning takes place within an authentic context. An authentic task, activity, or goal provides learning experiences as realistic as possible given the age and maturation level of the students and other environmental constraints, including safety and expense. The most important feature of this definition is to understand that "realistic experience" includes more than the situation; it includes both the context and the tasks that a learner performs (Honebein, Duffy & Fishman, 1993). Those tasks must be realistic in terms of the cognitive, physical, and social requirements. A realistic context includes as much fidelity as possible to what students will encounter outside school in terms of tools, complexity, and interactions with people (Williams & Dodge, 1992).

Authenticity is important to REALs for three reasons. First, it encourages students to take ownership of the situation and their own learning. Realistic problems hold more relevance to students needs and experiences, because they can relate what they are learning to problems and goals that they see every day. Second, it develops deeper and richer (indexicalized and conditioned) knowledge structures leading to a higher likelihood of transfer to novel situations. Finally, it encourages collaboration and negotiation. Complex problems require a team approach that provides natural opportunities for learners to test and refine their ideas and to help each other understand the content.

23.4.2.1. REAL Example: Anchored Instruction. One of the ways to create authentic instruction in a REAL is to anchor that instruction on a realistic event, problem, or theme (CTGV, 1990, 1992a, 1992b, 1993a, 1993d). Anchored instruction is fixed within a real-world event that is appealing and meaningful to students (Bransford et al., 1990). Anchored instruction involves complex contexts that require students to solve interconnected subproblems. Students share multiple perspectives, solutions, and processes (CTGV, 1992b). "At the heart of the model is an emphasis on the importance of creating an anchor or focus to generate interest and. enable students to identify and solve problems and pay attention to their own perception and comprehension of these problems" (Bransford et al., 1990, P. 123).

In anchored learning situations, students develop component skills and objectives in the context of meaningful, realistic problems and problem-solving activities. These complex contexts are called macrocontexts (Williams & Dodge, 1992, p. 373). The primary goal of anchored instruction (and REALs) is to overcome the problem of inert knowledge. For example, students in an instructional design and development class work in teams with actual clients to develop instruction that will be delivered to another group of students. They must define the problem, identify resources, set priorities, and explore alternative solutions-die same skills and abilities that are required during realistic, outside-of-the-classroom problem-solving and decision-making activities. This is in direct contrast to the way students develop component skills and objectives in a more traditional classroom environment by working simplified, compartmentalized, and decontextualized problems. Simply stated, it is the difference between providing meaningful, authentic learning activities and "I'm never going to use this" activities.

Anchored instruction shares many features of programs that are case based and problem based (Barrows, 1985; Spiro et al., 199 1; Williams & Dodge, 1992). The idea is to let learners experience the intellectual changes that experts feel when modifying their own understandings from working with realistic situations (CTGV, 1992b).

Effective anchors are intrinsically interesting, fostering ownership, and help students notice the features of problem situations that make particular actions relevant (Bransford et al., 1990, p. 123). The CTGV (1991) uses the following design principles when creating anchored instruction. First, they use a video-based presentation format because of the dramatic power of the medium and because of the use of multiple modalities, realistic imagery, and omnipresence in our culture. Second, they present a problem using actors and a narrative format for interest. Third, the problem solution requires a generative learning format in which students must identify pertinent information in the fourth feature, embedded data design. Fifth, the problem is complex with the possibility of multiple solutions and perspectives and requiring a team approach. Sixth, they use pairs of similar problems in different contexts to enrich the indexicalization of knowledge structures. Finally, they attempt to draw links across the curriculum to enhance the relevance of the problem.

There are several advantages to organizing curricula and learning around anchors and then progressing to hands-on projects (CTGV, 1993a). First, it is more practical and manageable for teachers to create anchors in the classroom than to try to arrange all of the resources, planning, and meetings around actual community-based projects. Second, the chance to work through one or more anchored problems Prepares students for actual problems they may undertake at a later time. Third, anchors provide a common experience and knowledge base that helps students share information with each other and with community members.

One of the CTGV's projects in anchored instruction is the Jasper Woodbury series (CTGV, 1992b). Jasper is a video-based series designed to promote problem posing, problem solving, reasoning, and effective communication. Each of Jasper's adventures is a 15- to 20-minute story in which the characters encounter a problem that the students in the classroom must solve before they are allowed to see how the movie characters solved the problem. The Jasper series helps students learn to break a problem into parts, generate subgoals, find and identify relevant information, generate and test hypotheses, and cooperate with others.

Information-rich learning environments are not designed as much as assembled, informally by individual learners. Hence they are constructivist in an almost literal sense of the term (Yacci, 1994, p. 1).

The third characteristic of REALs is that they are student centered. Student-centered learning environments place a major emphasis on developing intentional learning and lifelong learning skills. These skills include the abilities for self-reflection and metacognition.

23.4.3.1. Intentional Learning. Scardamalia and her colleagues noticed that passive or immature learners have certain characteristics that prevent them from becoming skillful problem solvers (1989). First, immature learners tend to organize their mental activities around topics rather than goals, promoting decontextualization and failing to see the relevance of the activity to their lives. Second, they tend to focus on surface features and do not examine a topic in depth. Third, they work straight ahead; that is, they tend to work until a task is finished. They do not take time to examine the quality of their work, nor do they make revisions in their work or thinking. Finally, they think of learning in an additive fashion rather than transforming and enriching their existing knowledge structures.

These characteristics are, in essence, the product of the conventional kind of schooling that we described earlier in the chapter. These behaviors prevent students from transferring their knowledge to new problems because they have learned in a decontextualized context and have learned with strategies that are decontextualized. They do not see the applicability of what they learned. Palincsar (1990, p. 37) states that:

To achieve transfer, it is necessary to attend to the context in which instruction and practice occur; transfer is likely to occur to the extent that there are common elements between the situation in which the children are learning this tactic and the situations in which such a tactic would be useful (p. 37).

Palincsar, Scardamalia, and Bereiter are leading proponents of the conviction that students must be taught to take more responsibility for their own learning to enhance the likelihood of transfer. They refer to this concept as intentional learning, or "those cognitive processes that have learning as a goal rather than an accidental outcome" (Bereiter & Scardamalia, 1989, p. 363). Palincsar and Klenk (1992) state that "Intentional learning, in contrast to incidental learning, is an achievement resulting from the learner's purposeful, effortful, self-regulated, and active engagement." To be intentional learners, students must learn to learn as well as learn to accrue knowledge. Learning to learn involves the teaching of generic skills as much as it does occupational or domain-specific skills. Teaching, too, takes on a revised role, for to teach for intentional learning means to cultivate those general abilities that facilitate lifelong learning (Palincsar, 1990). The main skills involved in teaching students to be more intentional are questioning, self-reflection, and metacognition, "or the awareness and ability to monitor and control one's activity as a learner" (Brown, Bransford, Ferrara & Campione, 1983, p. 212).

23.4.3.2. Questioning. Scardamalia and Bereiter (1991) believe that one of the first steps in developing intentional learners is by helping students take more executive control over what they decide to learn through the development of questioning skills. They point out that in a typical classroom, teachers ask the questions and that the question-generation and asking processes involve important high-level thinking skills and executive control decisions. Adults ask questions based on their needs, but teachers ask questions of students based on the teacher's perceptions of student needs. The students, therefore, do not ask questions related to their needs and do not learn to perform the analysis activities related to question generation. Research by Scardamalia and Bereiter (199 1) indicate that students can learn to ask questions to guide their knowledge building, thus assuming a "higher level of agency" and more ownership for their learning. In a student-centered REAL, students are given more executive control over their learning to enable them to take more ownership, to find more relevance and authenticity, and to learn lifelong-learning skills.

In addition to questioning, other intentional behaviors include goal setting, managing time, and setting priorities. Each of these helps students learn to manage their own learning and become more independent in the learning process. This independence leads to more ownership as the students discover that they are able to pursue their own needs and uncover information that is important to them.

23.4.3.2. Self-Reflection. A second skill in intentional learning is self-reflection. "Self-reflection implies observing and putting an interpretation on one's own actions, for instance, considering one's own intentions and motives as objects of thought" (Von Wright, 1992, p. 61). Von Wright writes that self-reflection involves the abstraction of meaning and is an interpretative process aimed at the understanding of reality. To understand the world in different ways involves modifying our conceptions of the world and our place in the world. It involves thinking about reality in alternative ways.

Von Wright (1992) goes on to describe two levels of reflection. One level of reflection has to do with the ability to reflect about features of the world in the sense of considering and comparing them in mind and thinking about on ways of coping in familiar contexts. This involves learning to think about implications and consequences of actions. A second level of reflection is the ability to think about one's self as an intentional subject of one's own actions and to consider the consequences and efficacy of those actions. This involves the ability to look at one's self in an objective way and to consider ways of changing to improve performance. The second level of reflection also involves metacognitive learning skills.

23.4.3.3. Metacognitive Skills. "Metacognitive skills refer to the steps that people take to regulate and modify the progress of their cognitive activity: to learn such skills is to acquire procedures which regulate cognitive processes" (Von Wright, 1992, p. 64). Metacognitive skills include taking conscious control of learning, planning and selecting strategies, monitoring the progress of learning, correcting errors, analyzing the effectiveness of learning strategies, and changing learning behaviors and strategies when necessary (Ridley, Schutz, Glanz & Weinstein, 1992). These abilities interact with developmental maturation and domain expertise. Immature learners can't to do this; they may have learned a single strategy, such as memorization, and then attempt to apply that to all situations.

Studies show that use of metacognitive strategies can increase learning skills and that independent use of these metacognitive strategies can be gradually developed in people (Biggs, 1985; Brown, 1978; Weinstein, Goetz & Alexander, 1988). Blakely and Spence (1990) describe several basic strategies for developing metacognitive behaviors:

1. Students should be asked to identify consciously what they "know" as opposed to "what they don't know."
2. Students should keep journals or logs in which they reflect on their learning processes, thinking about what works and what doesn't.
3. Students should manage their own time and resources, including estimating time requirements, organizing materials, and scheduling the procedures necessary to complete an activity
4. Students must participate in guided self-evaluation through individual conferences and checklists to help them focus on the drinking process.

23.4.3.4. REAL Example: Reciprocal Teaching. One of the manifestations of a REAL that emphasizes the development of intentional learning skills is reciprocal teaching. The context of reciprocal teaching is social, interactive, and holistic. Palincsar and Klenk (1992) used reciprocal teaching with at-risk first-grade students to develop reading skills. Palincsar and Klenk describe reciprocal teaching as:

... an instructional procedure that takes place in a collaborative learning group and features guided practice in the flexible application of four concrete strategies to the task of text comprehension: questioning, summarizing, clarifying, and predicting. The teacher and group of students take turns leading discussions regarding the content of the text they are jointly attempting to understand (Palincsar & Klenk, 1992, p. 213).

These strategies are the kinds of intentional learning strategies that encourage self-regulation and self-monitoring behaviors.

'The relationship of reciprocal teaching to REALs is founded on three theoretical principles based on the work of Vygotsky (1978) as described by Palincsar and Klenk (1992). The first principle states that the higher cognitive processes originate from social interactions. This is consistent with the constructivist theories described above. The second principle is Vygotsky's zone of proximal development (ZPD). Vygotsky (p. 86) described the ZPD as "the distance between the actual developmental level as determined by independent problem solving and the level of potential development as determined through problem solving under adult guidance, or in collaboration with more capable peers." Reciprocal teaching is designed to provide a zone of proximal development in which students, with the help of teachers and peers, take on greater responsibility for learning activities. Finally, Vygotsky's third principle advocates that learning take place in a contexualized, holistic activity that has relevance for the learners. In other words, we revisit the notion of authentic learning or anchored instruction.

How, then, does reciprocal teaching work? The process begins with a text that a class reads silently, orally, or read along, depending on the skill level of the students. Following each segment, a dialogue leader (students take turns) asks questions that deal with content or "wonderment" issues. The questions often stimulate further inquiry. The other students respond to the questions, raise their own

questions, and, in cases of disagreement or confusion, reread the text. The discussion leader is responsible for summarizing and synthesizing the reading and discussion and clarifying the purpose of the reading. The leader also generates and solicits predictions about the upcoming text to prepare for meaningful reading of the next segment. The teacher must model the appropriate behavior and provide scaffolding to sustain the discussion. (The preceding description is taken from Palinscar & Klenk, 1992.) The students, then, are involved in the higher-level thinking and decision-making activities usually within the realm of the teacher. With the help of the teacher, students share a zone of proximal development where they can learn questioning, summarizing, clarifying, and predicting activities so integral to metacognitive awareness.

Finally, why does reciprocal teaching work? Collins, Brown, and Holum (1991) posit the following reasons for its success (and, in a broader view, for the success of REALs):

1. The reciprocal teaching model engages students in activities that help them form a new conceptual model of the task of reading. They see reading as a process that involves reflection and prediction rather than just the. recitation of words. They learn to make what they are reading relevant to their needs and to monitor their progress and strive for clarification.
2. Teacher and student share a problem context while the teacher models expert strategies that the students learn to use independently.
3. Scaffolding is crucial in the success of reciprocal teaching. "Most importantly, it decomposes the task as necessary for the students to carry it out, thereby helping them to see how, in detail, to go about it" (Collins, Brown & Holum, 199, p. 11).
4. The students play both the roles of producer and critic. They learn cognitive activities that go beyond producing something for the teacher. They learn the self-monitoring activities and thinking processes involved in critiquing and improving their work.

The fourth characteristic of REALs acknowledges the transactional nature of knowledge and suggests that a shift be made to focus on social practice, meaning, and patterns (Roth, 1990). "All cooperative learning methods share the idea that students work together to learn and are responsible for one another's learning as well as their own" (Slavin, 1991, p. 73; see 35.3). Working in peer groups helps students refine their knowledge through argumentation, structured controversy, and reciprocal teaching. Additionally, students are more willing to take on the extra risk required to tackle complex, ill-structured, authentic problems when they have the support of others in the cooperative group. Cooperative learning and problem-solving groups also address students' needs for scaffolding during unfamiliar learning and problem-solving activities; therefore, with the support of others in the group, students are more likely to achieve goals they may not have been able to meet on their own. Constructivists argue that cooperative learning and problem-solving groups facilitate generative learning. Some of the generative activities that students engage in cooperative groups include (Brown et al., 1989):

1. Collective problem solving. Groups give rise synergistically to insights and solutions that would not come about individually.
2. Displaying multiple roles. Group participation means that the members must understand many different roles. They also may play different roles within the group to gain additional insights.
3. Confronting ineffective strategies and misconceptions. Teachers do not have enough time to hear what students are thinking or how they are thinking. Groups draw out, confront, and discuss both misconceptions and ineffective strategies.
4. Providing collaborative work skills. Students learn to work together in a give-and-take interaction rather than just dividing the workload.

Research indicates that cooperative learning, when implemented properly, is highly successful. Slavin (1991) provides the following four summary statements regarding research findings in cooperative learning:

1. Successful cooperative learning strategies always incorporate the two key elements of group goals and individual accountability.
2. When both group goals and individual accountability are used, achievement effects are consistently positive. His review found that 37 of 44 experimental/ control comparisons of at least 4 weeks' duration found significantly positive effects, with none favoring traditional methods.
3. Positive achievement effects are present to about the same degree across all grade levels (2-12), in all major subjects, and in urban, rural, and suburban schools. Effects are equally positive for high, average, and low achievers.
4. Positive effects of cooperative learning are consistently found on such diverse outcomes as self-esteem, intergroup relations, acceptance of academically handicapped students, attitudes toward school, and ability to work cooperatively.

23.4.4.1. REAL Example: Problem-Based Learning. Another manifestation of REALs is problem-based learning (PBL) (see 7.5, 20.3.4). PBL is "the learning that results from the process of working toward the understanding or resolution of a problem7' (Barrows & Tamblyn, 1980, p. 18). It found initial acceptance in the medical field and has grown to become a major learning system for a number of medical, law, and business schools. PBL reflects the REAL attribute that knowledge is constructed rather than received, for it is based on the assumption that knowledge arises from work with an authentic problem (Coltrane, 1993). Benor (1984) states that:

"Problem-based learning in the context of medical education means self-directed study by learners who seek out information pertinent to either a real-life or a simulated problem. The students have to understand the problem to the extent that its constituents can be identified and defined. The learners have then to collect, integrate, synthesize, and apply this information to the given problem, using strategies that will yield a solution" (p. 49).

How does problem-based learning work? We begin with Coltrane's description of the three fundamental theoretical principles of PBL:

1. Work on the problem begins with activating prior knowledge to enable students understand the structure of the new information, We also saw this principle used in the discussion about intentional learning that emphasized that students must ask themselves what they do know about a subject.
2. We also see the continual reference to the necessity for transfer in PBL, for when the learning context is similar to the situation in which the learning is to be applied, learning transfer is more likely to occur.
3. Learners must have opportunities to elaborate on the information presented at the time of learning in order to enhance their understanding. This is one of the main purposes for using cooperative learning strategies in REALs and was also seen as a part of reciprocal learning.

Savery and Duffy (1994) describe four characteristics of PBL. First, PBL environments include the learning goals of realistic problem-solving behavior, self-directed learning, content knowledge acquisition, and the development of metacognitive skills.

Second, Savery and Duffy state that PBLs are based on problems that are generated because they raise relevant concepts and principles that are authentic. Problems must be authentic because it is difficult to create artificial problems that maintain the complexity and dimensions of actual problems. Recall that we also encountered the need for complexity in the, REAL example of cognitive flexibility theory and anchored instruction. Realistic problems also have a motivational effect. They tend to engage learners more because they want to know the outcome of the problem.

Third, the actual presentation of the problem is a critical component of PBLs. Problems are encountered before any preparation or study has occurred (Barrows, 1980). The problem must be presented in a realistic way that encourages students to adopt and take ownership for the problem (Barrows, 1980; Savery & Duffy, 1994). The data must be embedded in the problem presentation (refer back to the example of anchored instruction) but must not highlight the critical factors in the case. Students must make their own decisions about what is critical and what is not because that is cognitively authentic: It reflects actual job performance (Savery & Duffy, 1994).

Fourth, the facilitator has a crucial role comparable to the roles described in anchored instruction and reciprocal teaching. The facilitator interacts with the students at a metacognitive level, helping them ask the right questions and monitor their own progress. Facilitators avoid expressing opinion, giving information, or leading to a correct answer. Their role is to challenge the students (from Savery & Duffy, 1994).

Cooperative learning is a critical component of PBL, for it is used from the beginning through the end of the problem-solution process. The group listens to the problem presentation together. They analyze the problem's components, recall' what they know, hypothesize, consider possible resources, and choose directions to go. They test and help each other. They work together on the solutions and reach consensus on final actions. The entire process from beginning to end is cooperative. Cooperative learning is also used for its motivational factors.

Problem discussion also increases motivation by gaining and maintaining student interest (attention), by relating the learning to student needs or helping students to meet personal goals (relevance), by providing conditions conducive to student success (confidence), and through the motivation provided by that mastery of the task(s) (satisfaction) (Coltrane, 1993, pp. 12, 13).

The PBL is the epitome of the REAL constructive learning process. Students work with problems in a manner that fosters reasoning and knowledge application appropriate to their levels of learning. In the process of working on the problem and with their peers, students identify areas of learning to guide their own individualized study. The skills and knowledge acquired by this study are applied back to the problem to evaluate the effectiveness of learning and to reinforce learning. The learning that has occurred in work with the problem and in individualized study is summarized and integrated into the student's existing knowledge structure.

The fifth requirement of REALs is that students engage in generative learning activities (see 3 1. 1. 1). People who learn through active involvement and use tools build an "increasingly rich implicit understanding of the world ..." (Brown et al., 1989, p. 33). Generative learning requires that students "engage in argumentation and reflection as they try to use and then refine their existing knowledge as they attempt to make sense of alternate points of view" (CTGV, 1993b, p. 16). Studies indicate that knowledge is more likely to be active and used when acquired in a problem-solving mode rather than in a factual-knowledge mode (Adams et al., 1988; Lockhart, Lamon & Gick, 1988). The concept of generative learning is an extension of the concept of constructing learning. Students cannot construct their own learning without generating something through active involvement.

Generative learning requires a shift in die traditional roles of students and instructors. Students become investigators, seekers, and problem solvers. Teachers become facilitators and guides, rather than presenters of knowledge. For example, rather than simply learning what objectives and goals are, students in a teacher education class generate lesson plans and objectives and then manipulate and revise them to solve new teaching problems. In generative learning, students apply the information they learn. Generative learning activities require students to take static information and generate fluid, flexible, usable knowledge. Generative learning, then, means that students are involved heavily with projects and creating solutions to authentic problems. A REAL model that relies heavily on projects is cognitive apprenticeship.

23.4.5.1. REAL Example: Cognitive Apprenticeship. Cognitive apprenticeship is modeled after the traditional apprenticeship (see 20.3) way of learning arts and crafts. It incorporates elements -of traditional apprenticeship and modem schooling. In apprenticeship, learners see products and processes of work. In traditional apprenticeship, the processes of an activity are visible and involve learning a physical and outwardly observable activity (Collins et al., 1991). The expert shows an apprentice how to perform a task, then watches and coaches as the apprentice practices portions of the task, and finally turns over more and more responsibility to the apprentice until the apprentice can perform the, task alone (Collins et al., 1991). Traditional apprenticeship deals with processes that are easily visible because they involve skills and producing products.

The goal of cognitive apprenticeship is to make processes that are normally invisible visible. In schooling, the process of thinking is usually invisible to both students and teachers. For example, the practices of problem solving, reading comprehension, and computation are not visible processes (Collins et al., 199 1). Brown, Collins, and Duguid (1989) point out that the term cognitive apprenticeship emphasizes that apprenticeship techniques can reach beyond observable physical skills to the kinds of cognitive skills associated with learning in schools. In a cognitive apprenticeship environment, the teacher attempts to make visible the thinking processes involved in performing a cognitive task. The teacher first models how to perform a cognitive task by thinking aloud. Then the teacher watches, coaches, and provides scaffolding as the students practice portions of the task. Finally, he or she turns over more and more responsibility to students and fades coaching and scaffolding until they can perform the task alone. "Cognitive apprenticeship supports learning in a domain by enabling students to acquire, develop, and use cognitive tools in authentic domain activity" (Brown et al., 1989, p. 39).

The differences between traditional and cognitive apprenticeship (Collins et al., 199 1) are important because they indicate where the effort must be placed on instruction design of learning activities. First, in traditional apprenticeship the task is easily observable. In cognitive apprenticeship, the thinking must be deliberately brought into the open by the teacher, and the teacher must help the students learn to bring their thinking into the open. Second, in traditional apprenticeship, the tasks come from the world, and learning is situated in the workplace. In cognitive apprenticeship, the challenge is to situate the abstract goals of school curriculum in contexts that make sense to students. Third, in traditional apprenticeship, the skills learned are inherent in the task. In schooling, students learn skills that are supposed to move across to different tasks. In cognitive apprenticeship, the challenge is to present a range of tasks to encourage reflection and to identify common transferable elements across tasks. The goal is to help students generalize and transfer their learning.

Cognitive apprenticeship and generative learning are closely linked, because the process of making cognitive processes visible means that students must create or generate things that represent those processes. Teachers must create work and tasks that represent the process of solving a problem, writing, or computation in addition to products. To examine the development of student thinking, an English teacher may ask for questions, themes, concept maps, and outlines before students begin writing. Math teachers are often notorious for telling students, "I want to see your work, not just the answer," so they can look for errors in the thinking process.

The elements of cognitive apprenticeship are present in the examples we have already examined. One of the purposes of reciprocal teaching, which we have already examined, is to teach children to perform some of the tasks of the teacher. Students watch the teacher model the tasks and then practice performing those tasks under the guidance of the teacher. In the Jasper series, students must demonstrate visible signs of the whole problem-solving process by asking questions, forming plans, finding resources, breaking a problem into its parts, and testing possible solutions with each other. In problem-based learning, the students work under the guidance of the teacher to solve real problems.

Generative learning is one of the simplest features of a REAL. It simply demands that students produce something of value. It is probably the most exciting part of a REAL, because students work on projects and tasks that are relevant to them and to their peers. It keeps students busy and happy while helping them learn. It also creates some unique assessment problems, which we will examine next.

The sixth and final REAL attribute is authentic assessment. Conventional schooling relies heavily on standardized and paper/pencil tests to measure the quantity of knowledge that students have accrued. Traditional tests, written reports, and grading schemes are inappropriate measures (Frederiksen & Collins, 1989), are time consuming to administer and score (Williams & Dodge, 1992), and are not always good indicators of how students will perform in actual problem-solving conditions. Williams also states that students are often assessed on skills different from ones that are taught and experience problem-solving assessments that tend to be subjective. Testing and assessment must recognize the importance of the organization of the knowledge base and its connectedness to contexts,

Wiggins (1989) contrasts authentic tests, which he describes as conceptualized, complex intellectual challenges with multiple-choice measures that he describes as fragmented and static. According to Wiggins, authentic tests include the following criteria:

1. The intellectual design features of tests and evaluation tasks must emphasize realistic complexity, stress depth more than breadth, include ill-structured tasks or problems, and require students to conceptualize content knowledge.
2. Standards of grading and scoring features should include complex multifaceted criteria that can be specified and that are reliable across multiple scorers. What constitutes a high level of performance should be explainable to students and teachers before they take the test. Teachers often claim that criteria are subjective; however this is seldom the case, for most criteria can be described with some thoughtful effort.
3. Tests and evaluations must be diverse and recognize the existence of multiple kinds of intelligences. In terms of fairness and equity, evaluations and assessments should allow students to use their strengths within areas that their interests lie.

23A.6.1. REAL Example: Learning in Design. Carver, Lehrer, Connell, and Erickson (1992) elaborate on this theme by proposing an extensive list of behaviors needed by students in a REAL. In their particular manifestation of a REAL, they consider the classroom a design community in which students design instruction for other students, documentaries for local media, and other exhibits for the community. Their program has the same goals of high-level thinking, reflection, and transfer as other REALs:

'The instructional virtues of these design experiences include the opportunity to develop and coordinate a variety of complex mental skills, such as decomposing a topic into subtopics, gathering data from a variety of sources, organizing diverse and often contradictory information, formulating a point of view, translating ideas into a presentation targeted at a particular audience, evaluating the design, and making revisions based on the evaluations (Carver et al., 1992, p. 386).

Again, there are several parallels with the other examples that we have discussed. Their REAL focuses on complex mental skills; analyzing, comparing, and manipulating information; working on authentic, community-based tasks; and working with others.

To fairly evaluate students working in this environment, teachers need a clear specification of skills students need for design tasks and prescriptions for how teachers can effectively support their skills (see also Agnew, Kellerman & Meyer, 1992). Specification of skills and prescriptions of support are two parts of assessment that must be linked for fair assessment. If a skill cannot be supported by. the teacher or some kind of scaffolding technique, then it cannot be fairly evaluated. It may, in fact, be outside the zone of proximal development and beyond the current capability of the student. One of the teacher's jobs in a REAL is to specify skills and performances that can be supported so that the student can grow in ability. Carver et al. (1992) break the important behaviors for their environment into:

1. Project management skills, including creating a timeline, allocating resources, and assigning team roles
2. Research skills, including determining the nature of the problem, posing questions, searching for information, developing new information, and analyzing and interpreting information
3. Organization and representation 'skills, including choosing the organization and structure of information; developing representations (text, audio, and graphics); arranging structure and sequence; and juggling constraints
4. Presentation skiffs, including transferring their design into media and catching and maintaining audience interest
5. Reflection skills, including evaluating the process and revising the design

Their criteria work for their learning environment. Other models may need to revise some of the specifics, though the five main categories provide an excellent starting place in specifying skills targeted for assessment. Goldman's (1994) discussion of assessment in the CTGV Jasper series suggests the following assessment areas: assessment of complex mathematical problem solving (math is the content domain of Jasper), measures of group problem-solving performance, assessment of extensions into other areas of the content area, and assessment of cross-curricular extensions. While conducting an assessment for a REAL is more work than conventional assessment, it is also an integral part of the learning process rather than a periodic quantifiable measure. Authentic assessment provides feedback and information that is useful for planning future learning.

23.4.6.2. Assessment of the REAL. There is yet another dimension to the issue of assessment: evaluation of the environment. To examine this, I'll look at another example of a REAL, case-based learning (CBL). "Case-based teaching exploits the basic capacity for students to learn from stories and the basic desire of teachers to tell stories that are indicative of their experiences" (Shank, 1990, p. 23 1). In case-based learning, teachers first teach students what they need to know to become interested in the case that will be examined. This, incidentally, is the prime difference between case-based learning and problem-based learning.

PBL problems differ from the typical case history in that they [PBL teachers] do not (initially) provide or synthesize all the information needed to solve the problem. In PBL, the problem is presented first, before students have learned basic science or clinical concepts, not after (Albanese & Mitchell, 1993, p. 53).

The information in CBL is presented to students in the form of stories. Students work with each other to analyze the stories or cases to abstract rules, heuristics, or practices that may be transferable to other cases. In CBL, the information is present within the case; students do not have to pursue other resources or individualized plans to learn more information as they do in PBL. Again, the main attributes of REALs are present: constructing knowledge, personal responsibility, cooperative learning, authentic context, and generative learning activities.

Williams (1992) developed a framework for comparing and evaluating methods of case-based instruction in the areas of teaching and learning and materials and curriculum. She poses 10 questions to guide the evaluation of case-based instruction (pp. 375-76):

1. Does instruction begin with a problem to be solved?
2. Does the teacher model expert problem solving in the context of a complex problem?
3. Are students given the opportunity to engage actively in solving problems, and does the teacher provide specific immediate feedback while students are solving problems?
4. What type of scaffolding is used to support students as they solve problems?
5. Does instruction emphasize metacognitive strategies as well as domain knowledge?
6. Are there frequent opportunities for both teacher and . students to assess how well learning is progressing? Is the type of assessment used appropriate for measuring the skills that are taught?

7. Are the problems authentic; that is, are they ones that would be solved by practitioners?
8. Are the problems realistically complex? Do their solutions involve multiple steps? Are the settings rich and detailed? Are multiple skills and concepts linked to each problem?
9. Are the problems presented in a way that makes complexity manageable, for example, using a story format, presenting them on video, and providing all relevant data?
10. Are problems sequenced to support students' needs at different stages of learning?

These questions apply to all REALs, not just case-based learning implementations. This common utility of these evaluation questions is another example of how much each of the examples of REALs that we have examined hold in common.

23.4.6.3. Evaluation Techniques. Assessment in REALs means that we have to consider more varied techniques. Neuman (1993), in her work with the Perseus hypermedia program, suggests several alternatives. First, she suggests that teachers use more observations, including evaluator observations of performance processes, think-alouds by students, and automatic transaction monitoring. Second, she suggests using interviews of students, instructors, and staff using both questionnaires and focus groups. Finally, she suggests the use of document and product analysis including assignments, syllabi, essays, journals, paths, reports, documentation, and presentations.

The issue of assessment is one of the more complex attributes of REALs because it is multidimensional. Assessment involves simultaneous assessment of both students and the environment. It also represents more than any other single attribute the depth of change necessary to implement a REAL. The move from paper/pencil tests to portfolio analysis, observations, interviews, and document analysis more than any single attribute signals the radical difference between REALs and conventional instruction.

We have looked at each of the six main characteristics of REALs: (1) constructivist heritage, (2) authentic instruction, (3) student responsibility, (4) collaborative learning, (5) generative learning activities, and (6) authentic assessment. Each characteristic of REALs builds on and uses the other. None is mutually exclusive; one is no more important than another: You cannot talk about one feature without incorporating the other. In effect, the characteristics of REALs mirror the comprehensive and integrated nature of REALs. The characteristics are symbiotic, with one feature both supporting and needing the others to create a successful, rich environment for active learning. In the next section, we examine some of the research related to the effectiveness of REALs.