12.3 EMERGING CONSTRUCTS AND LEARNING SYSTEMS

Whereas traditional approaches to computer-based learning have been rooted in behavioral learning principles, contemporary approaches are more often rooted in cognitive learning theories. They focus not on the product technology of the computer* but on the idea technologies afforded by the computer (Hooper & Rieber, 1995). Idea technologies tend to emphasize constructivist orientations to learning (see 24.3; Papert, 1981, 1993; Schwartz, Yerushalmy & Wilson, 1993; White, 1993).

Contemporary research with emerging technologies c . enters on the cognitive impact of people working in part nership with technology rather than studying the effects of technology on learning (see, for example, Perkins, 1985). This research can best be understood when classified as effects of versus with the computer on cognition (Salomon, Perkins & Globerson, 1991). Research on the effects of the computer on cognition attempts to determine if "cognitive residue" results as a consequence of the interaction between the individual and computer, such as an increase in general problem-solving ability or mathematical reasoning. Research with technology focuses on how human processing changes in distinct, qualitative ways when an individual is engaged in an intellectual activity using the computer as a tool. Taken interactively, an intellectual partnership is formed between the individual and the technology; the resulting changes to cognition cannot be understood when the individual or the technology are considered apart (see also 24.3).

A simple, yet profound, example of this intellectual partnership is evident in how human memory has been altered qualitatively with a relatively primitive technology: the pencil. The storage and retrieval strategies one uses to perform routine memorization are qualitatively different in cultures where writing instruments are universally available throughout one's life. Although computers are a long way from being as ubiquitous as pencil and paper, some researchers believe that computers offer a similar potential due to the unique processing support capabilities they provide. An emphasis on learning with media, as opposed to learning from media, may help to resolve some of the debate and controversy surrounding media research over the past 50 years [see, for example, Clark (1983) versus Kozma (1991b)].

This section focuses on learning systems and environments that have emerged based on contemporary psychological and pedagogical perspectives.

12.3.1 Psychological Constructs

12.3.1.1. Learning as the Active Construction of Knowledge. Most contemporary cognitive psychologists hold that learning consists of individual constructions of knowledge (see 7.4, 23.4.1.2). Learning is a personal event that results from sustained and meaningful engagement with one's environment (Bruner, 1961, 1985, 1986). This view also holds that learning cannot be viewed apart from the social and cultural contexts in which it occurs (Prawat & Floden, 1994). In education, the historical roots of constructivism are most heavily founded in developmental psychology and social learning theories. Educational practices encourage "equilibration" through the enabling processes of assimilation and accommodation described by Piaget (1952), and Vygotsky's (1978) construct of the zone of proximal development (Fowler, 1994). In some studies, researchers have even equated the computer as a scaffolding tool supporting readers as they enter, and maneuver within, the zone of proximal development (Salomon, Globerson & Guterman, 1989).

No single learning system design results from constructivist philosophies of learning [see Tobin (1993) for a collection of applications in science education]. Indeed, the notion that one person should design or plan a learning experience for another is antithetical to constructivism (Jonassen, 1991). Cognitive apprenticeships and anchored instruction, however, reflect ways in which instructional practice might change to accommodate the shift in learning from reception to construction (see Chapter 23). That these approaches lend themselves to widely varying interpretations is both an advantage and disadvantage. The literature is replete with innovative strategies for nurturing the construction process, but there are also many conflicts about the directions constructivist education should take (Strommen & Lincoln, 1992). Discovery-based, context-driven approaches are well suited to ill-structured domains, whereas traditional, directed instructional methods may be more appropriate in well-structured domains (Tripp, 1992).

12.3.1.2. Situated Cognition and Generative Learning. Two theoretical frameworks rooted in the principles advanced by Piaget and Vygotsky, situated cognition and generative learning (see 23.4.5, 31.1.1), are particularly promising for technologically enhanced learning environments (see, for example, Cognition and Technology Group at Vanderbilt, 1991). Situated cognition theory holds that learning should occur in environments that closely resemble those in which experts work. The most meaningful and useful kinds of learning are believed to be those embedded in activities that make deliberate use of social and physical contexts: "Activity, concept, and culture are interdependent. No one can be totally understood without the other two. Learning must involve all three" (Brown, Collins & Duguid, 1989, p. 33). Learning environments that support situated cognition closely resemble the apprenticeship system common in the craft professions (e.g., plumbing, carpentry, and tailoring) as well as professions that require extensive schooling (e.g., medical, dental, and legal). Likewise, learning about other subject matter in cognitive domains, such as mathematics, physics, and language arts, are often enabled through cognitive apprenticeships (Choi & Hannafin, 1995).

Current computing environments allow for a wide range of generative learning strategies to be incorporated into courseware (Jonassen, 1988). Generative learning models (see 31.3) suggest that meaningful learning results when the learner actively and consciously relates prior knowledge to new material and creates understandings based on these relationships (Wittrock, 1974, 1978; Wetzel, 1993). The role of instruction is to support activities and strategies that learners may use to generate meaning, and even to supply mechanisms if the learner is unable to do so on his or her own (such as in the case of very young children or novices). Generative learning requires learners to be proactive and mindful as they search for meaning by continually relating new information to what they already know. Generative activities include paraphrasing, summarizing, outlining, analytic reasoning, and mental imagery.

The work on computer-supported intentional learning environments (CSILE) by Marlene Scardamalia and her associates is one of the most extensive attempts to date to produce a computer system that facilitates generative learning (Scardamalia, Bereiter, McLean, Swallow & Woodruff, 1989). CSILE uses an instructional approach called procedural facilitation to support higher-order thinking skills. CSILE gives students tools and a context for making learning skills overt while providing just enough structure to enable cognitive processing, though such structure is reduced as students become more mature in their learning habits. CSILE permits students to enter information into a common database in a variety of representations (such as text, graphics, time-lines, etc.). As a shared database, students are able to enter newly discovered information while at the same time are prompted to describe points of confusion and enter questions they might have. Students are also encouraged to evaluate and refine each other's entries. CSILE makes students responsible for their learning while providing them with tools and strategies that support metacognitive processes, such as goal setting, elaboration, and self-evaluation. Early research with CSILE has produced very promising, though largely anecdotal, results.

12.3.1.3. Microworlds. A microworld represents the simplest case of a domain that is still recognizable by an expert in the domain. As a learner becomes more proficient, the microworld can become more complex and sophisticated. Most importantly, the learner decides how to structure and direct the microworld according to individual needs.

Microworlds tend to focus on conceptual and qualitative understanding within relatively narrow and limited sets of interrelated constructs. This is an especially important departure from traditionally quantitative domains, such as mathematics and physics. Traditional instruction often enables students to calculate accurately the mathematics of motion problems without changing student's fundamental conceptions or "personal theories" of physical science (diSessa, 1988, 1993; Roschelle, 1991; Roschelle & Greeno, 1987). Carefully designed physics microworlds, many with gamelike features, have been found to qualitatively alter students' conceptions of physics problems (White, 1984,1992,1993).

Computer-based microworlds provide opportunities for students to explore and experience phenomena intuitively and formulate hypotheses that may run counter to intuition (diSessa, 1982). Using well-conceived n-microworlds, children as young as sixth-graders demonstrate better qualitative understanding of physics, and comparable calculation capabilities, as their twelfth-grade counterparts (White & Fredericksen, 1987, as cited in Striley, 1988). The aims of such systems are often fundamentally different from traditional, objectivist aims; the methods required for their study must reflect these basic differences.

12.3.2 Emerging Learning Systems

Concrete applications of constructivism to educational practice have become increasingly prevalent [see, for example, STELLA described by Steed (1992)]. Some have detailed ways to alter dramatically classroom practices from instruction to construction using traditional media and materials (see Fosnot, 1989). However, several important developmental projects have resulted that utilize the computational and processing power of computers to create, or alter, learning environments. In this section, we describe four noteworthy systems, each of which has been studied extensively, that utilize technology in innovative ways to support the construction on personal understanding: Logo MicroWorlds, Jasper Woodbury, Voyage of the Mimi, and Citizen Kane.

12.3.2.1. Logo MicroWorlds. Early research on Logo yielded conflicting results and interpretations, though the most consistently positive results were in the affective domain (Clements, 1985). Children using Logo often reported positive attitudes, were enthusiastic about learning, and improved in - their motivation to cooperate in groups (e.g., Burnet & Higginson, 1984; Hawkins, Sheingold, Gearheart & Berger, 1982).

Among the most contentious issues was whether Logo influenced general cognitive skills, such as problem solving (see 24.5.1). Several notable studies indicated no improvements in problem-solving from using Logo (Pea, 1987; Pea & Kurland, 1984), whereas others reported successes (Clements & Gullo, 1984; Kynigos, 1993; Resnick, 1990; Soloway, Lockhead & Clement, 1982). The source of these discrepancies likely were rooted in the different research approaches. For example, Clements and Gullo (1984) studied the effects of Logo programming over a wide range of issues including cognitive style, metacognitive ability, and cognitive development. Nine children were provided Logo programming experience, while nine other children were provided traditional computer-related instructional activities. The Logo children showed increases in metacognition, whereas the control group did not. In contrast, Pea and Kurland (1984) tested the learners' capacity to solve specific related problems rather than general ones, such as problem planning. Consistent with the bulk of research on generalized problem solving, Logo failed to yield favorable results, suggesting that improvements, if they occur, may be of an exceedingly broad nature.

Inconsistent transfer of cognitive skills may also be attributable to different approaches to studying transfer (see, for example, Lehrer & Littlefield, 1993). According to Salomon and Perkins (1989), many unsuccessful Logo studies were concerned with low-road transfer. This is not surprising, since young students rarely attain very sophisticated programming skill with Logo, even after 2 years (Leron, 1985). Successful studies, such as reported by Clements and Gullo (1985), emphasized high-road transfer. Students received guidance and structure as they worked with Logo as compared to the free discovery methods often misattributed to Papert.

Papert (1987) contended that early research on Logo effects was fundamentally misguided, since Logo, like the constructivist philosophy on which it is based, must be viewed holistically. Context, culture, and activity must be viewed together; reducing the experience to its individual parts necessarily changes the nature of the experience. Logo, Papert asserts, is but one knowledge tool in the student's culture. Hypotheses related to the effects of Logo on learners are akin to hypotheses related to the effects of hammers and wood on building good houses.

Recent research with current generations of Logo, such as MicroWorlds Project Builder, MicroWorlds Language Art, and MicroWorlds Math Links, focuses on a special interpretation of constructivism that Papert calls constructionism. Papert chose the term constructionism because of the emphasis placed on building artifacts of understanding, such as children building their own games to learn about fractions (Harel, 1990, 1991; Harel & Papert, 1991; Kafai, 1994). Each MicroWorlds system extends underlying Logo vocabulary, syntax, and primitives with a variety of tools, color palettes, shapes, animation, fonts, and other features to make use more readily apparent and intuitive. The constructionist philosophy reflected within each system remains essentially the same as the early versions of Logo. The means for empowering learners, however, evolved to make transparent the tool uses of the computer in constructing artifacts of their understanding.

12.3.2.2. Jasper Woodbury. The Jasper project (see also 23.5.1.1) is primarily video based, featuring both videotape and videodisc versions, though recent efforts have extended the materials to other multimedia contexts (Marsh & Kumar, 1992). Though traditional computer technology is not central to the project, the importance of microprocessor technology in accessing segments in the video scenarios, rapidly and precisely, is integral to its design. The Jasper series represents an application of constructive approaches to learning, specifically generative learning and anchored instruction (Cognition and Technology Group at Vanderbilt, 1992a, 1992b).

The Jasper series uses brief vignettes to present a complex problem to be solved. The episodes use real-life drama to present a "macro context" for students to engage in problem solving. The purpose of the macro context is to anchor instruction, and situate cognition, in a context that is perceived as real and meaningful (Choi & Hannafin, 1995; Eylon & Linn, 1988). The video presents the problem context or goal, such as how to transport a wounded eagle from a remote site as quickly as possible under a number of situational constraints. All information. necessary to solve the problem is embedded in the video, which students can play repeatedly to search for, and extract, relevant information. However, students are not prompted in explicit ways to the embedded information, nor are they told how to solve the problem. Instead, they must generate their own tentative solutions, test them, and revise their approach and theory accordingly (see 23.5.1).

In addition, analogs and extensions are available to challenge students to transfer their problem-solving skills to other contexts and to make flexible their knowledge representations. Analog problems change various parameters of the original video episode, such as the wind conditions on the day of the eagle's rescue. Extension problems challenge students to integrate their knowledge of the problem in the video to other curriculum areas, such as Lindbergh's flight across the Atlantic.

Students using the Jasper Series demonstrated significant gains in storywriting, vocabulary usage, and acquisition of relevant subject matter. In addition, they demonstrated better transfer of complex problem solving when provided with the anchored instruction. Those given the anchored instruction scored equally well on standardized achievement tests, despite the fact that the Jasper series was not designed to promote the kind of learning that these tests typically measure (Cognition and Technology Group at Vanderbilt, 1993,1994).

12.3.2.3. The Voyage of the Mimi. A similar use of video to anchor, or situate, learning is found in The Voyage of the Mimi, a multimedia curriculum package developed at the Bank Street College of Education. Mimi integrates print, video, and computer materials in learning about science and mathematics. Video is used to present a realistic, fictional account of the adventures of the crew of the Mimi, a ship hired by a team of scientists to study humpback whales (additional Mimi voyages have also been produced using contexts such as Mayan archaeology). This context is used throughout the Mimi curriculum to teach about problem solving in mathematics and science. The Mimi video shows scientists at work using mathematics and science as vital tools for understanding and helping whales. 'Me video is likewise used to invite the viewer to become involved in the same knowledge through the interactive computer materials.

The Mimi's computer materials provide students with a wide variety of interactive activities, such as simulations and games, that parallel the adventures of the Mimi's crew. The computer materials are organized around four modules: maps and navigation, ecosystems, whales and environments, and introduction to computing. The printed materials provide paper-and-pencil problems and activities for students to complete. The Mimi materials were developed to be sufficiently flexible to provide teachers with multiple levels of entry to the materials in their classrooms (Martin, 1987). Teachers can select to use some or all of the materials to augment or replace an existing curriculum. The materials and activities support a wide range of learning outcomes and can be used individually or in cooperative groups.

12.3.2.4. Citizen Kane. Citizen Kane is a multimedia learning system adapted by Rand Spiro and his colleagues from Orson Wells's classic movie of the same title. It was an effort to create an environment within which understanding of complex concepts could be accomplished flexibly. Cognitive flexibility theory (see 23.4.1.3) addresses the problem of learning simple versus complex knowledge by having students use interactive multimedia (i.e., computer-controlled videodisc using hypertext principles) to reveal multifaceted issues (Spiro & Jehng, 1990). Cognitive flexibility theory describes the knowledge structures that learners need to interpret and analyze complex and novel situations. Certain domains, such as literary criticism, appear to demand cognitively flexible learners. The first half of Citizen Kane introduces the basic characters and plot. The final half of the original movie was edited to allow various scenes to be juxtaposed. These scenes were theme related and provided students a model and environment with which to manipulate and understand the complex nature of the characters and the movie.

12.3.3 Reflections on Emerging Learning Systems

Contemporary applications of technologies have shifted the teaching-learning paradigm dramatically. There has been unprecedented attention to the student's role in the instructional process. Indeed, even the term instruction is considered a pejorative to some in describing emerging learning systems. Whereas designers have traditionally focused on what subject matter to teach and how best to teach it (e.g., presentation and sequencing strategies), emerging learning systems place greater responsibility on the learner. As a result, many long-standing notions about teaching, learning, and design have changed.

The role of the computer has changed from a transmitter of knowledge to a tool that aids in the construction of knowledge (see, for example, Forman & Pufall, 1988). Motivational considerations have also shifted, though not always for the better. Some designers emphasize high-interest activities in order to arouse students, often to the detriment of the learning activity (see, for example, Coleman, Koballa & Crawley, 1992).

The shift away from a subject-matter focus and conventional design wisdom has not been without debate and detractors. Diverse interpretations and views exist as to what such transitions mean to instructional technology (see, for example, Duffy & Jonassen, 1992; Kember & Murphy, 1990; Perkins, 1992). While R&D has thrived, there has been comparatively little impact in the "real world." School curriculum is still largely focused on basic skills and formal subject-matter information [see Raghavan & Glaser (1995) for an interesting alternative]. Many still feel that the computer's role is to support, or even duplicate, accepted instructional practices. To many, increased focus on learning processes over outcomes is seen as inappropriate and ill advised.. However, recent examples, such as CSILE and Jasper, provide well-rooted, empirically grounded systems and clearer distinctions and benefits between the use of technology to promote learning of fixed knowledge versus learner empowerment. Unfortunately, strongly rooted efforts remain the exception, not the rule.

The learning systems we presented depart from traditional educational technology-instructional design applications in three major ways. First , they represent a shift away from traditional notions of learning from media to more contemporary notions of learning with media. The tools and resources described here present a fundamentally different approach to teaching and learning. For example, many tools and resources support direct and active learner manipulation and organization of information in order promote deeper or different understandings. Second, the aforementioned systems are firmly rooted in contemporary research and theory on teaching and learning and focus on the construction of knowledge rather than the transmission of information. For example, microworlds not only assist students in solving problems but also help learners think about, experience, and manipulate problems in order to develop expertise, modify knowledge, or change personal beliefs (Resnick, 1991). Third, the functional features support, extend, and enhance the processes of teaching, learning, and thinking. 111-structured domains, of which there are many, require tools that support varied perspectives, encourage the student to analyze from a variety of points of view, and avoid artificial simplification of complex concepts for the sake of expedience. Easier is not necessarily better.

This review not only illustrates the potential of emerging technologies to redefine many of our traditional notions about teaching, learning, and technology, but it punctuates the need for researchers to continue to develop, assess, and evaluate these new constructs and systems. Research is needed to understand better how to optimize the capabilities of learners and technologies, as well as to conceptualize how emerging technologies can be utilized to improve classrooms and schools.