For the past two decades, a plethora of national reports have indicated that American public school students have been lagging behind their international counterparts in such areas as mathematics, science and technology (National Commission of Excellence in Education, 1983; National Research Council, Board on Mathematical Sciences, 1989; The United States Congress Office of Technology Assessment, 1995). Since the early eighties, these reports have prompted the examination of our schools’ curricula and fueled the drive for school reform (Knapp & Glenn, 1996). The contention of these and a host of other studies is that schools need to shift their focus from transmitting information to students to enabling them to seek information, process it and thereby construct their own knowledge (Knapp & Glenn).

A growing recognition on the part of educators of the inappropriateness of traditional views of teachers as all-knowing disseminators of knowledge and students as passive recipients has resulted in new ways of thinking about the process of teaching and learning. New models which focus on the use of multimedia and communications technologies as tools to assist teachers in helping students take responsibility for their own learning, become knowledge explorers and collaborate with others to solve real problems are a natural outgrowth of a new emphasis on engaged and active learning. (Garrett & Weiner, 1998; Barker, 1999).

Knapp and Glenn (1996) present such a model aimed at examining differences between transmissive (traditional) schools and restructured (Constructivist) ones. Central to their model is a paradigm shift which focuses on students learning to construct their own knowledge by engaging in “activities that require them to think critically, solve problems, and seek answers to their own questions” (Knapp and Glenn, p. 7). To accomplish this shift, the curriculum and classroom learning environment are restructured to facilitate student inquiry, research, and problem solving activities. This shift in emphasis also requires a concomitant shift in assessment techniques. In their model, traditional short answer and essay tests are supplanted by authentic situations in which students are required to demonstrate their ability to express, apply and defend their knowledge and skills. Technology also plays a pivotal role in the configuration of a Constructivist school. Knapp and Glenn call for the integration of technologies which support “research, analysis, problem-solving, and communication” (Knapp and Glenn, p. 7). Many researchers (Beal, 1995; Johnson, 1994; Liu, 1998) believe that, used in this context, hypermedia development can help students construct knowledge, develop higher order thinking skills and, possibly, promote problem-solving skills.

As new models of learning have been developed, it has been recognized that learners act as active constructors of knowledge (Duffy and Jonassen, 1992). Within this constructivist framework, which is concerned with the process of how we construct meanings of our world as well as the results of the constructive process, cognitive tools can help learners organize, restructure and represent what they know. Jonassen and Reeves (1996) have summarized the foundations of cognitive tools research and have identified the following key principles in the context of hypermedia design:

• Cognitive tools have their greatest effectiveness when they are applied to constructivist learning environments.
  
  
• Cognitive tools empower learners to design their own representations of knowledge rather than absorbing knowledge representations preconceived by others.
  
  
• Ideally, tasks or problems for the application of cognitive tools should be situated in realistic contexts with results that are personally meaningful to learners. (p. 698).
  
  

Of central concern for educators is how to best utilize these cognitive tools in a way that will help their students develop the ability to think for themselves, apply concepts that they learn and analyze and evaluate the material that they study. In a nutshell, students need to be able to manipulate various technologies in a way that will enable them to synthesize information, construct knowledge and ultimately learn how to think for themselves. Montgomery (1999) suggests that by implementing an innovative approach based on SURWEB’s media shows and learning, these goals can be positively addressed through the use of student created hypermedia presentations.

SURWEB (http://www.surweb.org), a federally funded project aimed at providing resources to students and teachers in remote, rural areas, provides application technologies which allow students to manipulate Internet based multimedia. In addition to acquiring a large number of pictures and text, SURWEB has also developed a media presentation tool that enables users to easily access and organize all forms of digital information and then to display their products in a simple, yet professional manner. The SURWEB presentation tool allows users to: 1) visit a site of interest and view the relevant information available on-line; 2) select images and text to make a media presentation utilizing multimedia from the database of images, texts and sounds; and 3) choose to import or upload images, sounds and text from any source in order to customize and enhance the media presentation. The software application has been extensively field tested and upgraded according to formative evaluation.

The most often used SURWEB hypermedia tool feature is the import site, where students and teachers can select data from other sites on the Internet to build classroom projects. In the past five years, more than 7,000 user-created projects of this type have been created and stored for later use on the SURWEB server. The use of SURWEB hypermedia tools encourages knowledge creation through the ability to research, access, evaluate, organize, and display information that is relevant to the content and performance standards for the classroom. In addition, the tools are useful for peer coaching, collaborative learning, and the accommodation of a greater variety of student orientations to learning.

This action research study will explore the use of SURWEB media shows as cognitive tools for creating interactive learning environments in a seventh grade social studies classroom.

This study focuses on attempts by sixth grade social studies students to develop hypermedia products which fall into the knowledge construction category of Nelson and Palumbo’s (1992) cognitive framework. The premise of this study is that, by using hypermedia in this manner, students might engage in higher levels of cognitive skills including analysis, synthesis, and evaluation and develop creative thinking and research process skills. In other words, by using SURWEB to develop their own hypermedia programs, students will be engaged in a process which requires them to externalize the constructs of their own knowledge.

The purpose of this study was to investigate the effects of SURWEB Media Show construction on the development of complex knowledge structures, creative thinking and research process skills of sixth grade social studies students. For the purpose of the study Media Shows were defined as web-based programs created through the use of SURWEB software which enable students to link animation, video, graphics, sound and text together in meaningful ways. Hypermedia, as used in this study refers to documents which seamlessly incorporate text, graphics, sound, video and hyperlinks (links to other documents). Reinking (1994) points out that hypermedia “provides a means to express ourselves in ways that reflect more directly the complexity of our thinking and the inter-relatedness of ideas.”

Yankovich, Meyrowitz and van Dam (1985) observe that the use of hypermedia “allows authors and groups of authors to link information together, create paths through a corpus of related material, annotate existing texts and create notes that point readers either to bibliographic data or to the body of reference material.” Hypermedia can thus allow students and teachers to create an automated encyclopedia which enables readers to browse through linked, cross-referenced, annotated text (hypertext) in an orderly but non-sequential manner.

This study incorporates both traditional and non-traditional assessment measures to determine how students’ cognitive structures and thinking skills are affected when they develop hypermedia programs. The research questions which guide the study include:

1. What is the effect of SURWEB hypermedia development on students’ ability to construct complex knowledge structures?
   
   
2. What is the effect of SURWEB hypermedia development on students’ acquisition of information problem solving skills?
   
   
3. What is the effect of SURWEB hypermedia development on students’ creative thinking skills.
   

Hypermedia and the Construction of Knowledge

Papert (1987) and Nicol (1989) and a growing number of educational technologists support a “constructivist” school of hypermedia which is based on the belief that students should reconstruct their own understanding of their studies and assume a more active role in their education. Marchionini (1988) maintains that when students create their own hypermedia stacks, they present knowledge in a way that matches their own schematic framework of understanding. Ashton (1992) has suggested that when students reconstruct knowledge for themselves they tend to learn more because the information is internalized. Bodner (1986) argued that until recently the accepted model for instruction was based on the hidden assumption that knowledge can be transferred intact from the mind of the teacher to the mind of the learner. He points out that teaching and learning are not synonymous; teachers may teach and teach well, but students will not necessarily learn. Von Glaserfeld (1979) and other Constructivists advocate an approach in which students identify topics or issues, locate resources, plan investigations and activities, and practice self-evaluation (all with teacher support). Under this model the emphasis is shifted from activities that teachers do, to those that students perform during the learning process.


Hypermedia Development and Cognition

A considerable body of research exists which supports the usefulness of hypermedia-based learning environments (Conklin, 1987; Hammond, 1989; Heller, 1990; Jonassen, 1996). The concept has existed for quite some time as indicated by an early reference by Nelson (1967) who defined the process as “a combination of natural language text with the computer’s capacity for interactive branching and the dynamic display of nonlinear (information).” According to Jonassen, a hypermedia system consists of fragments of information that can be text, graphics, sound or video. He refers to these fragments of information as “nodes” and notes that nodes can be linked to other nodes to create complex knowledge structures. Jonassen considers hypermedia knowledge structures to be similar to processes through which information is stored in the human memory system. A user of a hypermedia system can navigate through the knowledge structure and determine which path to follow as well as the sequence and length of viewing. Structures are created so that the user can “jump around” within the program to related or more detailed information as desired. Much of the early literature on hypermedia, is descriptive or theoretical and addresses users of hypermedia systems rather than the effect of engaging in the construction of hypermedia on students’ knowledge structures.

Taylor (1980), in his pioneer work on computers in education, described three major functions of computer usage: computers as tutors, tutees and tools. When used as tutors and tutees, computers help students understand difficult concepts and check student comprehension of those concepts. Used as a tool, computers empower students to create and manipulate knowledge, organizing it in a variety of cognitively useful ways. Toolistic use of the computer has gained increasing acceptance among educators with the advent of inexpensive and easy to use hypermedia authoring software (Hasselbring, Goin, & Wissick, 1989). Fabris (1992) found that the use of a such software provides rich learning experiences for all students, regardless of cultural, social, economic, or ethnic background.

Most recent studies on hypermedia have focused on this “toolistic” approach in which teachers create their own multimedia stacks for students to use as an alternative to traditional texts (Jonassen, 1986; Bowers & Tsai, 1990). This type of multimedia implementation requires more knowledge on the part of the classroom teacher since he/she is no longer merely “applying” multimedia but “creating” multimedia presentations and then “applying” them to course content in meaningful ways (Landow, 1989).

Other studies have focused on students as hypermedia authors (Brigham, Hendricks, Kutcka, & Schuetta, 1994). Hooley and Toomey (1993) advocate the student use of computers to store, share, and reconstruct knowledge through hypermedia applications. Pea (1991) has suggested that the full educational potential of multimedia technology will not be realized until students are empowered to create their own multimedia projects to use them to communicate their understanding and organization of the topic being studied. Pea’s claims are supported by Stevens (1993) who studied the use of computer software programs and student activities utilizing the computer in secondary social studies classrooms. Stevens concluded that the successful use of computers in social studies instruction depends upon the design of student created projects.

Lehrer (1993) constructed a framework for the cognitive processes involved in the development of hypermedia which identifies the four types of cognitive processes involved. These processes include planning, transforming, evaluating and revising the hypermedia document. With Erickson and Connell (1994) he studied the cognitive effects of the development of hypermedia projects on ninth grade students in an American History class and concluded that the students acquired skills such as interpreting information and communicating knowledge through the development of intricate hypermedia documents.

Lund and Hildreth (1997) investigated an instructional model that incorporated the personal computer and a multimedia authoring tool into an assignment to write and illustrate a multimedia story. Subjects were 21 public school students in a fifth-grade classroom. Students were instructed in the writing of interactive stories using the process approach to writing and the software and a scoring rubric was used to evaluate the finished stories. The researchers concluded that: (1) students’ writing improved in quantity and quality; (2) students were more motivated to complete the assignment; (3) the computer allowed for greater flexibility in writing due to its nature as a word processing tool; (4) even a single computer could be effectively used in integrating computer technologies into the classroom, and; (5) “there were inherent difficulties associated with integrating computer technologies into classrooms---many students did not have experience with computers, even those who do have experience may lack keyboarding skills, and finding time to accomplish such a project requires outside help.” These findings suggest that, with appropriate support, computers can be used to enhance writing instruction and improve the quality of written work.

A number of studies have reported benefits when hypermedia authoring tools are placed in the hands of students. Velasco and Mendivil (1992) reported that when students use authoring tools in a constructivist environment, motivation increases. In addition, the use of graphics, sounds, video clips, and peripheral devices for photographic material is very attractive for image-centered learning, and the hypermedia environment demands careful planning which implies the exercise of higher skills in structural design and thinking.

Reed and Rosenbluth (1992) focused on students as creators rather than users of programs to determine whether collaboratively creating HyperCard stacks that presented information on four decades (i.e. the 1920s, 1930s, 1945-60 and the 1960s) affected the amount of knowledge and the inter relatedness of informational units students had on the assigned decades. Thirteen high school seniors utilized hypermedia authoring tools to produce a multimedia-based strand of instruction during a four week academic enrichment program. Results indicated that students engaging in the development of the hypermedia program featuring the various factors affecting the values of certain decades exhibited an increased awareness of the inter relatedness of these factors and that these changes in perceptions were due to the linking nature of the hypermedia authoring language.

Oughten and Clevenger (1997) examined the effects of prior computer experiences on graduate students’ cognitive levels as they developed Hyperstudio projects. Students participating in their study were asked to describe their use of Hyperlogo, HyperStudio’s programming language and to create concept maps which depicted the inter-relationships of the language’s features and processes. The researchers found that students with more experience focused on the procedural aspects of the Hyperlogo while students with less experience concentrated on the declarative aspects of the language.

Heidenreich (1997) investigated the effects of using Hyperstudio, a hypermedia authoring tool, and concluded that the Hyperstudio method of instruction was not effective in raising the achievement level of participating seventh grade social studies students. However, he noted at students who used the authoring tool had a more positive attitude toward learning, their partners, and cooperative learning in general.

The positive effects of hypermedia on student perceptions were also noted by Smith (1993) who investigated the hypermedia learning environment from the perspective of high school students learning about U.S. culture. Qualitative methods were used to determine student perceptions about learning with hypermedia tools. Data were collected from daily observations, interviews, student products, formal and informal discussions and from the use of questionnaires. The researcher concluded that there are a number of student perceptions that are supportive of the use of a hypermedia learning environment. Students indicated that the nature of hypermedia products promoted considerable motivation. Hypermedia documents were viewed by students as much more public than traditional work. Producing a hypermedia document promoted understanding at a deeper level than using a traditional assignment. Teachers participating in the study concurred that learning was enhanced and all participants in the program indicated the use of a hypermedia learning environment to be advantageous.

In a similar descriptive experiment involving Constructivist use of hypermedia Volker (1992) hypothesized that students would more readily use primary source information if their goal was to create a hypermedia project. During the three-month study students created the program design and content treatment for portions of a math and science hypermedia project while teachers served as content advisors. Prior to the implementation and field test of the project, participants completed attitudinal surveys on technology, their fear of it, their level of knowledge before using the materials, and their preference for working on their own.

Following the field test, 35 student users, 3 teachers, and 3 student producers again responded to attitudinal survey instruments. Participants reported that, compared to traditional instruction, hypermedia was more interesting and that they preferred using it instead of traditional textbooks, videotapes, etc. Teachers also expressed enthusiasm for the motivational aspects of this approach, claiming that students showed more interest in math and science. Nearly all teachers recognized and supported the shift in emphasis from teacher centered instruction to student based learning.

As a follow-up to a four year longitudinal study, Tierney and Kieffer (1997) examined the impact of hypertext on students’ learning in science and English classes. The first study compared the impact of HyperCard stacks and regular textbook presentation of ideas in biology on ninth grade students assigned to study these materials. Students in the second study developed their own multimedia projects in science and English courses using HyperCard and their standard textbooks. The researchers analyzed interviews, videotaped observations, and outcome and process measures and concluded that hypermedia was an effective alternative to traditional print-based texts. Specific findings indicated that (1) students suggested that the hypertext offered alternatives to standard print texts, (2) students considered hypermedia projects more interesting than conventional texts and assignments, (3) hypermedia texts created by students contained more main ideas, enlisted more illustrations, were more multilayered, and provided clearer links between illustration and the text, (4) hypertext allowed a flexible exploration of ideas across several layers simultaneously, (5) students were motivated to explore its possibilities and were willing to share new findings with other students and (6) the use of hypertext supported the acquisition of complex ideas and relations between those ideas involving science tasks. Findings suggest that students view the advantages of hypertext as providing a way to “architecture” a space that affords different opportunities for engagement by others. Tierney concluded that students engaged in developing hypermedia “developed notable experience in problem solving, integrating visual representations of ideas with text, and developing refining and restructuring ideas.”

A considerable body of evidence suggests that when employed as a constructivist tool, the use of hypermedia authoring software by students can significantly increase their ability to retain knowledge. Classic research by Geisman (1988) suggests that students retain 20 percent of what they see, 40 percent of what they see and hear, and 70 percent of what they see, hear and do. Since the use of student produced hypermedia projects facilitate seeing, hearing and doing by students, it is reasonable to assume that students’ achievement on tests and retention of information will likely improve when these approaches to learning are utilized by teachers in constructive ways.


Hypermedia Construction and Information Problem Solving

Many educators and researchers advocate engaging students in activities which require more than the simple application of rules and procedures. Collins, Brown and Newman (1989) contend that few educational resources, including interactive multimedia, are devoted to higher-order problem solving activities.
To date, there appears to be little research into whether creative thinking, problem-solving and other higher order thinking skills can be developed through the use of hypermedia.(Herrington and Oliver, 1998).

Polya (1957) developed a model for problem-solving which included four principal processes:

1. Understanding the problem
   
   
2. Developing a plan
   
   
3. Implementing and carrying out the plan
   
   
4. “Looking back” or accessing the effectiveness of the plan

In the first step, understanding the problem, the problem solver represents the problem, identifies the unknowns of the data, and determines the conditions of the problem. The second step of the model consists of developing a plan to solve the problem which meets the conditions identified in step one. In this step, the problem solver determines the relationships between the data and the conditions of the problem in order to formulate a plan. In the third step, the problem solver implements and executes a plan designed to address the conditions of the problem. The final step of the model is assessment or looking back. In this step, the problem solver determines the “correctness” of the solution and assesses whether it adequately meets the conditions of the problem.

Lehrer’s (1993) four-step framework for describing the cognitive processes involved hypermedia development are in close alignment with Polya’s problem-solving steps. Lehrer’s planning stage can be seen as being equivalent to Polya’s first steps of understanding the problem and devising a plan. According to Lehrer, planning tasks include the determining the audience, content, topics, relationships among topics, and interface design of the hypermedia project. Lehrer’s transforming stage is similar to Polya’s second step of implementing and carrying out the plan. Transforming tasks include accessing, interpreting, and determining how to represent the information. Lehrer’s evaluation and revision stages are equivalent to Polya’s looking back step. Evaluation and revision tasks include assessing the presentation and organization of information, testing the program soliciting the feedback of users, and, finally, restructuring the program to make it more accessible or meaningful.

Reed and Liu (1994) compared the effects of programming in BASIC versus developing hypermedia with HyperCard on graduate students’ problem solving skills. They found that the BASIC group increased their problem-solving skills, whereas, the HyperCard group did not. Liu (1998) conducted a follow-up study to the Reed and Liu (1994) study. The treatment in the Liu study specifically focused on using HyperCard’s scripting features. Findings indicated that the students increased their problem-solving skills from pretreatment to post treatment.

McKenzie (1995) developed a research model and rubric for assessing students’ information problem-solving skills based on the work of Lehrer and several models including standards and rubrics proposed by NCREL. McKenzie’s research cycle includes seven steps: questioning, planning, gathering, sorting and sifting, evaluating and reporting. According to McKenzie, the student passes through several repetitions of the first six stages until sufficient information is gathered to form insights worthy of reporting. Such research is based upon a decision to be made or a solution to be proposed--- research devoted to the exploration of essential questions. During the research process, students maintain “research logs” which track the reasoning used as well as the research actions taken while cycling through the process.


Hypermedia Construction and Creative Thinking

The term “creative thinking” as used in this study refers to a set of generalized mental abilities, identified by Torrance (1984) commonly presumed to be brought into play in creative achievements. Other researchers (Wallach & Kogan, 1965; Guilford, 1970; Hattie, 1980) have used the term in various contexts to refer to such constructs as divergent thinking, productive thinking, inventive thinking or imagination. Over a period of 25 years Torrance and his associates have developed several batteries of test activities at predicting the ability of an individual to behave creatively. These batteries are incorporated into the Torrance Tests of Creative Thinking (Torrance & Ball, 1984) which approach the construct of creativity through the theoretical rationale of a series of criterion referenced indicators which include fluency, originality, abstractness of titles and resistance to premature closure.

Harkow (1996) used a combination of strategies to improve creative thinking skills in second- and third-grade gifted students. Sixteen students were targeted for the intervention. Over a 12-week implementation period, students participated in 90-minute interventions twice weekly. The intervention was comprised of 30-minute creative problem-solving encounters with peers; 30 minutes of computer software use to produce original writing, and to experiment and create in open-ended settings; and a 30-minute period of activities alternating between relaxation and imagery exercises and the use of imagery in creative writing. There were four objectives to the intervention: (1) increasing verbal and figural creativity; (2) increasing figural and verbal fluency; (3) increasing figural and verbal originality; and (4) increasing verbal flexibility. The Torrance Tests of Creative Thinking, Figural and Verbal Models and the Inventory of Creative Behaviors were used to assess the impact of the intervention. Informal teacher observations were conducted throughout implementation.

The Inventory of Creative Behaviors was completed weekly and at the conclusion of the implementation period. Findings indicated that the proposed number of students met the projected percentage of increase of 80 percent or above in overall figural and verbal creativity, verbal originality, and verbal flexibility.

Liu (1998) examined whether engaging elementary school students in hypermedia authoring would promote their creative thinking. The researcher reported significant correlations between hypermedia development and scores on the Torrance Tests of Creative Thinking.

In summary, theoretical and empirical research by cognitive psychologists suggests that higher order thinking and problem solving are complex human activities. Researchers have demonstrated that cognition involves processing incoming external information and encoding this information within memory. There are many types of knowledge structures or schemata that we rely on for thinking and these structures are constantly being modified based on new information that is received and processed. Learning can be defined as the process of modifying existing knowledge structures to accommodate new information for specific purposes. Solving problems involves understanding and representing the problem and then applying different knowledge structures to develop strategies to determine a solution.

When used appropriately, student produced hypermedia projects have the potential to enhance student learning and subsequent achievement, as well as cognitive and social skill development (Carver, 1992; Wisnudel, 1994). This study will specifically examine SURWEB hypermedia construction tools and the effect these tools can have upon the construction of knowledge, creative thinking and information problem solving. Is is hypothesized that the use of SURWEB’s web-based authoring environment by students will produce instructional benefits very similar to those facilitated by Hyperstudio and other software-based hypermedia authoring tools.

While few studies have specifically focused on sixth grade social studies and hypermedia, a considerable body of research suggests that the use of hypermedia authoring tools such as HyperCard and Hyperstudio (applications whose functions are similar to the hypermedia capabilities of SURWEB media tools) have had positive effects on the ability of students to construct knowledge and solve problems. Therefore, it was hypothesized that sixth grade social studies students who create their own Media Shows will achieve statistically significant higher ratings on performance based measures of complex knowledge structures and information problem solving. It was also hypothesized that students actively engaged in hypermedia construction would score higher than their traditionally instructed peers on a norm referenced measure of creative thinking.

Participants

The participants in this study were 260 sixth grade social studies students selected from a rural Southern Utah area intermediate school. Students were described by their teachers as being primarily average to slightly above average in ability. The sample was drawn from the classrooms of ten social studies teachers who volunteered to participate in the study. Forty students were randomly assigned to a control group and the remaining students were randomly grouped into five pods of forty to reduce the influence of individual teacher input upon the outcomes being assessed.


Setting for the Study

Students assigned to the control groups met in traditional seventh grade social studies classrooms. Students were seated in rows typical of most sixth grade classrooms but were allowed to move their desks together to work in small groups to complete assignments with partners.

Students in the experimental groups met in computer labs. All labs contained Macintosh computers which were networked and connected to the Internet via a Local Area Network. An LCD projector or large screen monitor was available for use by the teacher and students for presentations.

Instruction for both groups was conducted during normally scheduled times throughout the school day and both groups spent approximately six weeks engaged in content of the lessons. Instructional methods for the two groups were identical except for the use of the SURWEB media show construction tool by students in the experimental groups.

In the school participating in the study, computer labs were frequently used by math, science, English, social studies and other classes. Because of this widespread use by students, it is unlikely that the novelty of the environment exerted either a positive or a negative effect on the performance of students in the experimental group. In a follow up survey conducted as part of the study, virtually all students reported that they felt comfortable in the computer lab and considered it a part of the normal, everyday school environment.


Assessments

A number of researchers have pointed out that specialized assessment techniques are required to analyze student thinking and problem solving processes (Larkin and Rainard, 1984; Schoenfeld, 1985; Marshall, 1995). Marshall claims that the emphasis should be placed on addressing three critical questions: 1) What is it that individuals learn?,” “How do they learn and remember it?,” and “Can they use it again?” (Marshall, pp. 171-172). Jonassen (1996) points out that traditional measures, such as multiple choice tests are limited because critical thinking and problem-solving are complex skills which are difficult to assess because they are context dependent. Several researchers (Oughton & Reed, 1998; Reed & Rosenbluth, 1992; Liu and Pederson, 1998) advocate the use of concept maps and metacognitive learner logs and problem sets to explore these constructs.

Assessment of Knowledge Structures

A cognitive representation or concept map can be viewed as a structural representation of a learner’s knowledge in a subject domain (Ruiz-Primo & Shavelson, 1995). Jonassen (1993) describes concept maps as “the spatial representation of ideas and their inter-relationships that are stored in memory” (p.3). Jonassen, Beissner and Yacci (1993) claim that “concept maps generated by learners at different points in a curriculum may be used to identify learners’ progression in the attainment of differentiated and organized structural knowledge”(p. 157). The intended use of concept maps is to “tap into a learner’s cognitive structure and to externalize, for both the learner and the teacher to see, what the learner knows” (Novak and Gowin, 1984, p. 40).

To assess the impact of SURWEB hypermedia construction on the development of complex knowledge structures, students participating in the study were asked to produce concept maps of their knowledge of states or foreign countries during a six week social studies unit. As an indicator of cognitive growth, concept maps constructed at the beginning of the activity were compared to maps constructed at the end of the unit.
Assessment of the concept maps was based on an adaptation of Novak and Gowin’s (1984) scoring system. This scoring system is outlined below.

• 1 point for each example of specific and valid processes or concepts
  
  
• 1 point for each relationship or link between two concepts
  
  
• 2 points for each labeled link or relationship
  
  
• 3 points for each valid cross link that did not illustrate a synthesis between concepts
  
  
• 5 points for each valid level or segment of the schemata
  
  
• 10 points for each valid cross link which showed a meaningful connection between one segment of the schemata and another segment.
  

The concept maps were used to determine, in part, whether the structural knowledge of the students’ mental models became more sophisticated over the five-week period.

Student projects and research questions from the learner logs were also reviewed by evaluators who assessed their content using the six general categories of Bloom’s (1956) Cognitive Taxonomy. To simplify the process of scoring and to insure a higher degree of reliability, Bloom’s six categories (Knowledge, Comprehension, Application, Analysis, Synthesis and Evaluation) were compressed into three rankings: Lower Level Thinking (Knowledge and Comprehension), Medium Level Thinking ( Application and Analysis) and Higher Level (Synthesis and Evaluation). Three evaluators were employed to code the project and learner log data. Scores of one two and three were awarded according to the cognitive levels represented by the data set. If the ratings of the first two evaluators were not identical, the data sets were scored by a third evaluator. In cases where the first three evaluators’ codings were not at identical categories, a fourth evaluator coded the set until agreement was reached. Reliability of the scores was calculated using the codings of all evaluators. The inter-rater reliability coefficient was .90.

These dependent measures were used to answer the first research question: What is the effect of SURWEB hypermedia development on students’ ability to construct complex knowledge structures?


Assessing Information Problem Solving

Jonassen (1996) claimed that having students document their problem solutions encourages them to “focus on, classify and evaluate the steps and strategies they used to solve a problem” (p. 277). By implementing this method of assessment, patterns of skill acquisition and types of metacognitive strategies that students employ can be monitored (Jonassen).

Learner logs which included concept maps of student research procedures and descriptions of information problem solving processes were used to examine the impact of SURWEB hypermedia construction on the development of problem solving skills. Logs entries were collected at the beginning and end of the unit and comparisons used to assess student growth in information problem solving skill.

Students maintained learner logs throughout the study in which they listed research questions and recorded their progress toward solving information problems. At the end of the six week unit students were asked to produce a research processes concept map and to describe the stages involved in conducting their research. Students’ information problem-solving skills were assessed using a seven step rubric proposed by McKenzie (1995). Details of the rubric appear below.

1. Questioning: A researcher recognizes decisions, issues and problems when exploring a topic.

• 5 points: Discovers independently an issue or problem which needs a decision or solution after exploring a topic
  
  
• 3 points: Formulates questions about topics with adult help to evaluate the question to focus on issues and problems
  
  
• 1 point: Relies upon adults to state questions and topics

2. Planning: A researcher identifies sources of information likely to build understanding.

• 5 points: Selects high quality sources independently and efficiently
  
  
• 3 points: Selects sources with mixed success
  
  
• 1 point: Wanders from source to source without questioning which
  source will be most helpful in answering research questions

3. Gathering: A researcher collects and stores information for later consideration.

• 5 points:Collects and organizes important information for retrieval independently
  
  
• 3 points: Collects information with some degree of organization
  
  
• 1 point: Loses track of most important information

4. Sorting: A researcher reorganizes information so that the most valuable becomes readily available to support understanding.

• 5 points: Creates structure which provides a coherent and clear focus
  
  
• 3 points: Creates partial organization of information
  
  
• 1 point: Leaves information as gathered

5. Synthesizing: A researcher recombines information to develop conclusions and decisions.

• 5 points: Creates an original conclusion or decision based on information
  
  
• 3 points: Reorganizes and combines conclusions and decisions of others
  
  
• 1 point: Restates the conclusions and solutions of others

6. Evaluating: A researcher determines whether the information gathered is sufficient to support a conclusion.

• 5 points: Tests solutions and decisions to see if supporting information is adequate
  
  
• 3 points: Looks for missing information
  
  
• 1 point: Reaches a hasty conclusion

7. Reporting: A researcher translates findings into a persuasive, instructive, or effective product(s).

• 5 points: Creates and presents an original product which effectively addresses original problem or issues
  
  
• 3 points: Provides a product which offers some insight with regard to the original problem or issues
  
  
• 1 point: Shares only the work or thoughts of others

Assessors reviewed the research process component of students’ learner logs and evaluated their final projects using the above rubric. This dependent measure was used to answer research question two: What is the effect of SURWEB hypermedia development on students’ acquisition of information problem solving skills?


Assessment of Creative Thinking

The Torrance Test of Creative Thinking (TTCT), a norm referenced test of creative thinking ability was used to assess the impact of SURWEB hypermedia construction on the development of creative thinking skills. The TTCT was administered to students in both control and experimental pods and results compared to assess differences in performance between the two groups. Several scoring reliability studies indicate that reliability coefficients above .90 for the TTCT can be achieved if scorers are properly trained in the use of the scoring guidelines which accompany the tests. To insure reliability of the assessment, each test was independently scored by two evaluators. In cases where discrepancies between scores for each of the five scales were identified, tests were rescored by a third evaluator. This method yielded an inter-rater reliability of .93.

The content and construct validity of the TTCT has been explored in a factor analytic study by Mourad (1976), a comparative study by Rungsinan (1977), and in a developmental study by Alieldin (1982). These studies assembled considerable theoretical and empirical research to support the content validity of each of the indicators used in the battery.

Two major predictive validity studies have also been conducted (Torrance, 1980; Torrance & Wu, 1981). These studies published in 1981, involved young adults tested at the time they were in elementary school and followed up in 1980, twelve years later. Three criteria of creative behavior were used: the number of publicly recognized creative achievements reported, the quality ratings of five judges of the three most creative achievements described by the subjects and the creativity of their future career images. Pearson product-moment coefficients of correlation were obtained for each of the three separate criteria with each of the creativity indicators. Measures of Fluency, Originality and Elaboration yielded a validity coefficient of .51, a value statistically significant at the .01 level.

Scores obtained through the use of the TTCT were used to compare the creative thinking skills of students in the control and experimental groups at the end of the 5-week unit. This dependent measure was used to answer research question three: What is the effect of SURWEB hypermedia development on students’ creative thinking skills.


Experimental Design

Students and teachers were selected to participate in this quasi experimental action research study which employed a pretest/post test control group experimental design (see Table 1). Subjects in both groups were pretested prior to the treatment and post tested following completion of an instructional unit on states and foreign nations.

The school participating in the study reported high attendance rates during the period of time in which data were collected. As a control for subject mortality, only students who completed all assessments were included in the analysis of data. Subject mortality posed no threat to the validity of the study in that 93 percent of the subjects included in the original sample concluded the unit and completed all assessments. With the exception of a few daily absences, the consistency of the control and experimental groups remained unchanged.

For the purpose of the study, the presence or absence of SURWEB Media Show construction as an instructional strategy was the independent variable. Student performance on the three cognitive assessment instruments (TTCT, concept maps and information problem solving rubrics) were the dependent variables. The research design is detailed in the table 1 which appears below.

Group	Assignment	n	Treatment	Pretest	Post test
C1	Control	40	Regular Instruction	Concept Map Research Rubric	Concept map Research Rubric TTCT Project Rubric
E1	Experimental	40	SURWEB Instruction	Concept Map Research Rubric	Concept map Research Rubric TTCT Project Rubric
E2	Experimental	40	SURWEB Instruction	Concept Map Research Rubric	Concept map Research Rubric TTCT Project Rubric
E3	Experimental	40	SURWEB Instruction	Concept Map Research Rubric	Concept map Research Rubric TTCT Project Rubric
E4	Experimental	40	SURWEB Instruction	Concept Map Research Rubric	Concept map Research Rubric TTCT Project Rubric
E5	Experimental	40	SURWEB Instruction	Concept Map Research Rubric	Concept map Research Rubric TTCT Project Rubric

Procedures

In April and May of 2000, 260 sixth grade social studies students in ten classrooms were randomly selected to one of two groups or pods. One pod of forty students was randomly selected to be the control group while the remaining five pods comprised the experimental group.

Two volunteer teachers taught a unit on U.S. states and foreign nations to students in the control group using project-based instruction. Students in the control group utilized traditional library resources and classroom materials to research and prepare a report on either a state or a foreign country.

The remaining eight regular program social studies teachers taught the same unit to students assigned to the experimental group. The primary difference between the two groups was the use of SURWEB Media Show construction tools by students of the experimental group. Instead of the traditional library and classroom materials, students in the experimental group researched the topic and prepared an on-line report using the SURWEB tools.

Both groups were instructed by their regular program teachers. Both groups used the same textbooks and were given equal amounts of time to research the content of the unit and to produce a report on their findings. Learning objectives and assessments for both groups were identical

Due to the limited number of computers available in the labs, students worked in small groups. In order to prevent the student grouping variable from influencing the results of the study, students in the control group also worked in small groups to complete their projects and learning tasks.

At the beginning of the unit, both groups were given pretest assessments which included the construction of concept maps of prior knowledge, completion of an initial learner log entry and a measure of information problem-solving skills. Upon completion of the geography unit, all students were completed alternative form post test assessments.

In addition to the pretest/post test data, demographic and descriptive data were collected using the SURWEB Student Technology Survey questionnaire designed to provide information regarding students’ level of experience with technology. Data generated by the questionnaire were used to explore relationships between levels of experience, gender, and other factors which might influence the dependent variables.

At the beginning of the study, students assigned to the two groups were given a pretest in order to obtain benchmark data and to check initial group equivalence. Scores on the pretest measures for the control and experimental groups were compared using a two-tail t-test for paired samples. This statistic can be used to determine if significant differences exist between the means of two samples. The test makes three assumptions about the scores obtained in causal-comparative research. The first assumption is that scores form an interval or ratio scale of measurement. The second is that scores in the population under study are normally distributed. The third assumption is that score variances for the population under study are equal. Borg (1989) maintains that even if the assumptions underlying the t-test are violated, the t-test will still provide, in most instances, an accurate estimate of the significance level for differences between sample means. Analysis of the data obtained by using the pretest/post test instruments suggest that the three assumptions were met. Examination of the means and a paired samples t-test indicated no statistically significant difference between the two groups (Table 2). This finding supports the assumption that there were no initial differences between the ability levels of the two groups.

Concept Maps (First Assessment)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CLOG1RP	40	.259	14.7000	4.214	.666
E1LOG1RP	40	.259	14.6250	3.985	.630

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
.0750	4.994	.790	.09	39	.925

95% CI (-1.522, 1.672)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CLOG1RP	40	.205	14.7000	4.214	.666
E2LOG1RP	40	.205	14.5250	3.870	.612

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
.1750	5.103	.807	.22	39	.829

95% CI (-1.457, 1.807)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CLOG1RP	40	.306	14.7000	4.214	.666
E3LOG1RP	40	.306	15.1500	3.431	.542

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-.4500	4.546	.719	-.63	39	.535

95% CI (-1.904, 1.004)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CLOG1RP	40	-.001	14.7000	4.214	.666
E4LOG1RP	40	-.001	15.0250	3.355	.530

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-.3250	5.389	.852	-.38	39	.705

95% CI (-2.049, 1.399)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CLOG1RP	40	.075	14.7000	4.214	.666
E5LOG1RP	40	.306	14.9000	3.153	.498

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-.2000	5.070	.802	-.25	39	.804

95% CI (-1.821, 1.421)


Concept Maps (First Assessment)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CMAP1	40	-.076	37.1750	14.875	2.352
E1MAP1	40	-.076	36.7350	13.799	2.182

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
.4500	21.046	3.328	.14	39	.893

95% CI CI (-6.281, 7.181)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CMAP1	40	.250	37.1750	14.875	2.352
E2MAP1	40	-.076	36.4500	14.102	2.230

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
.7250	17.756	2.807	.26	39	.798

95% CI (-4.954, 6.404)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CMAP1	40	.060	37.1750	14.875	2.352
E3MAP1	40	.060	38.4250	14.753	3.123

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
.-1.2500	24.008	3.796	-.33	39	.744

95% CI (-8.928, 6.428)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CMAP1	40	.053	37.1750	14.875	2.352
E4MAP1	40	.053	37.9500	14.158	2.239

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-.7750	19.990	3.161	-.25	39	.808

95% CI (-7.168, 5.618)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CMAP1	40	-.163	37.1750	14.875	2.352
E5MAP1	40	-.163	41.8250	18.198	2.877

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-4.6500	25.313	4.002	-1.16	39	.252

95% CI (-12.745, 3.445)

A t-test for paired samples was used to compare the performance of the two groups on the post tests administered at the end of the unit. This statistical technique was deemed appropriate because the assumptions required for the use of the parametric t-test were met, e.g. the data were interval.

To answer research question one: What is the effect of SURWEB hypermedia development on students’ ability to construct complex knowledge structures?, a paired samples t-test was used to compare the means of the experimental and control groups on two assessments: 1) holistic scoring of the learner log research questions and the completed project using a scoring model based on the categories of Bloom’s Taxonomy, and 2) concept map scores assigned using an adaptation of a rubric developed by Novak and Gowin (1984).

To answer research question two: What is the effect of SURWEB hypermedia development on students’ acquisition of information problem solving skills?, a paired samples t-test was used to compare the performance of students in the control and experimental groups on information problem solving measures included in the learner log. Student problem-solving concept maps and research process descriptions were scored using a seven step rubric proposed by McKenzie (1995).

To answer research question three: What is the effect of SURWEB hypermedia development on students’ creative thinking skills?, scores on the Torrance Test of Creative Thinking were compared using a paired samples t-test to compare the creative thinking abilities of students in the control and experimental groups

Upon completion of the unit, post test scores of the two groups on the various assessments were compared using the statistical methods described above. As Tables 3 through 7 indicate, there were significant differences between the performance of the control and some of the experimental groups participating in the study.

Table 3 compares the assessment scores of the control and experimental groups on the final concept map included in the learner log. This map provided indication of the complexity of knowledge structures developed by students in both groups during the five week unit. Four of the five experimental groups scored significantly higher on the concept map assessment than their control group counterparts. Although not statistically significant, the mean of experimental group three was higher than the control group’s and approached significance at the .05 level.

On the basis of the data outlined in Table 3, it was concluded that the use of SURWEB hypermedia development tools by students resulted in an increased ability to construct complex knowledge structures. This conclusion supports the findings of Tierney and Kieffer (1997) who reported that the use of hypertext supported the acquisition of complex ideas and relations between those ideas. Tierney's conclusion that students engaged in producing hypermedia "developed notable experience in problem solving, integrating visual representations of ideas with text, and developing, refining and restructuring ideas" was also supported.

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CMAP2	40	-.010	43.4250	18.628	2.945
E1MAP2	40	-.010	55.2750	18.799	2.972

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-11.8500	26.593	4.205	-2.82	39	.008

95% CI (-20.355, -3.345)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CMAP2	40	.107	43.4250	18.628	2.945
E2MAP2	40	.107	57.5250	22.534	3.563

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-14.1000	27.659	4.373	-3.22	39	.003

95% CI (-22.946, -5.254)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CMAP2	40	.022	43.4250	18.628	2.945
E3MAP2	40	.022	51.4500	23.880	3.776

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-8.0250	29.963	4.738	-1.69	39	.098

95% CI (-17.608, 1.558)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CMAP2	40	-.062	43.4250	18.628	2.945
E4MAP2	40	-.062	53.7000	21.369	3.379

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-10.2750	29.206	4.618	-2.23	39	.032

95% CI (-19.615, -.935)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CMAP2	40	-.002	43.4250	18.628	2.945
E5MAP2	40	-.002	58.3500	23.655	3.740

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-14.9250	30.144	4.766	-3.13	39	.003

95% CI (-24.565, -5.285)

Holistic scoring of the learner log research questions and the completed project using a scoring model based on the categories of Bloom’s Taxonomy provided additional evidence supporting the positive effects of the SURWEB media construction tool on the development of complex knowledge structures. Student projects and research questions from the learner logs were scored using a three point scale based on the six general categories of Bloom's (1956) Cognitive Taxonomy. To simplify the process of scoring and to insure a higher degree of reliability, Bloom's six categories (Knowledge, Comprehension, Application, Analysis, Synthesis and Evaluation) were compressed into three rankings: Lower Level Thinking (Knowledge and Comprehension), Medium Level Thinking ( Application and Analysis) and Higher Level (Synthesis and Evaluation). Table 4 compares the cognitive level rankings of learner log research questions and final projects developed by students included in the control and experimental groups.

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CLOG2RP	40	.194	17.2000	3.532	.558
E1LOG2RP	40	.194	18.7750	2.983	.472

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-1.5750	4.156	.657	-2.40	39	.021

95% CI (-2.904, -.246)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CLOG2RP	40	.181	17.2000	3.532	.558
E2LOG2RP	40	.181	17.8250	3.257	.515

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-.6250	4.348	.688	-.91	39	.369

95% CI (-2.016, .766)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CLOG2RP	40	.265	17.2000	3.532	.558
E3LOG2RP	40	.265	18.5000	2.298	.363

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-1.3000	3.667	.580	-2.24	39	.031

95% CI (-2.473, -.127)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CLOG2RP	40	.153	17.2000	3.532	.558
E4LOG2RP	40	.153	18.4000	2.687	.425

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-1.2000	4.096	.648	-1.85	39	.071

95% CI (-2.510, .110)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CLOG2RP	40	.083	17.2000	3.532	.558
E5LOG2RP	40	.083	18.8500	2.370	.375

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-1.6500	4.086	.646	-2.55	39	.015

95% CI (-2.957, -.343)


Learner logs which included concept maps of student research procedures and descriptions of information problem solving processes were used to examine the impact of SURWEB hypermedia construction on the development of information problem solving skills. Three of the five experimental group scored significantly higher on the assessment of information problem solving skills than the control group. Though not statistically significant, the mean scores of groups two and four were slightly higher than those of the control group. The comparatively low performance of group two may have resulted from lack of familiarity with concept maps on the part of several students in the sample. Evaluators noted that several students in this group failed to complete the task of concept map construction. Though inter-rater reliability was relatively high (.81) the scoring rubric used to assign individual scores is not without a degree of subjectivity. It is also possible that subtle variations in scoring may have resulted in the failure to detect statistically significant differences in group four’s performance. On the basis of this analysis, it was concluded that SURWEB hypermedia construction has a positive impact on the development of information problem solving skills by students.

This finding supports the conclusions of Liu (1998) and Reed and Liu (1994) that hypermedia development by students leads to an increase in problem-solving skills. Tierney and Kieffer (1997) also reported that students engaged in hypermedia development exhibited gains in problem-solving ability.

Table 6 compares the performance of control and experimental groups on the Torrance Test of Creative Thinking. Three of the five experimental groups scored significantly higher than the control group on this measure. The remaining two groups produced means that approached significance at the .05 level.

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CTTCT	40	-.212	82.5050	10.175	1.609
E1TTCT	40	-.212	87.4000	9.179	1.451

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-4.8950	15.080	2.384	-2.05	39	.047

95% CI (-9.718, -.072)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CTTCT	40	-.089	82.5050	10.175	1.609
E2TTCT	40	-.089	87.3675	9.053	1.431

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-4.8625	14.211	2.247	-2.16	39	.037

95% CI (-9.407, -.318)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CTTCT	40	.118	82.5050	10.175	1.609
E3TTCT	40	.118	86.7475	10392	1.643

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-4.2425	13.658	2.160	-1.96	39	.057

95% CI (-8.611, .126)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CTTCT	40	-.113	82.5050	10.175	1.609
E4TTCT	40	-.113	89.1350	8.582	1.357

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-6.6300	14.032	2.219	-2.99	39	.005

95% CI (-11.118, -2.142)

Variable	Number of Pairs	Correlation	Mean	SD	SE of Mean
CTTCT	40	.176	82.5050	10.175	1.609
E5TTCT	40	.176	86.6500	12.423	1.964

Paired Differences

Mean	SD	SE of Mean	t-value	df	2-tail Sig.
-4.1450	14.611	2.310	-1.79	39	.081

95% CI (-8.818. .528)


Analysis of scores on the Torrance Test of Creative Thinking support the conclusion that the creative thinking abilities of students who use the SURWEB media tool are enhanced. This conclusion is supported by the findings of Liu (1998) and Harkow (1996) who reported correlations between hypermedia development and increased scores on tests of creative thinking.


Limitations

It would be difficult to generalize the results of this study to all classrooms due to the limitations of the research design. Future research should address the application of the SURWEB media construction tool in other curricular areas and explore the issue of how hypermedia construction can best be utilized in broader classroom contexts. In addition, issues of individual learning preferences and gender differences should also be addressed. While further research is called for, the potential of the SURWEB approach to enhance the development of complex knowledge, problem-solving and creative thinking appears promising. When used in a constructivist, learner-centered classroom environment, SURWEB can be a powerful tool for facilitating higher order thinking skills.

Aiken, M.C. (Ed.). (1992). Media use in education. The Encyclopedia of Educational Research, 2, 805-813

Ali-el-din, M.T. Torrance indicators of creative thinking: A developmental study (Doctoral dissertation, University of Georgia, 1978). Dissertation Abstracts International, 1979, 39 (7-A), 4129. (University Microfilms No 70-10, 151).

Ashton, P.T. Editorial. Journal of Teacher Education, 43 (5), 322.

Barker, Bruce O. (1999). SURWEB: A Visual Literacy Tool Which Promotes Engaged Learning for Teachers and Students. Paper presented at the Annual National Conference on Creating the Quality School (Memphis, TN, March 26-28, 1999).

Beal, J.W. (1995). The Civil War: An example of integrating Linkway into the social studies curriculum. Hyper Nexus 5, 3-6.

Bodner, G.M. (1986). Constructivism: A theory of knowledge. History of Chemical Education, 63 (10), 220-229.

Boone, R., & Higgins, K. (1992) Hypermedia applications for content area study guides. Reading and Writing Quarterly: Overcoming Learning Difficulties, 8, 379-393.

Borg, Walter R. and Gall, Meredith D. (1971). Educational Research: An Introduction.
New York: David McKay and Company. 303-307.

Bowers, D. & Tsai, C. (1990) HyperCard in educational research: An introduction and case study. Educational Technology, 19-24.

Brigham, F.J., Hendricks, P.L., Kutcka, S.M. & Schuette, E.E. Hypermedia supports for student learning (Report No. IR 017 028). Indianapolis, Council for Exceptional Children. (ERIC Document Reproduction Service No. ED 378 965).

Carver, S. (1992). Learning by hypermedia design: Issues of assessment and implementation. Educational Psychologist, 27 (3), 380-400

Collins, A., Brown, J.S., & Newman, S.E. (1989). Cognitive apprenticeship: Teaching the crafts of reading, writing and mathematics. In L.B. Resnick(Ed.). Knowing, Learning and Instruction: Essays in Honour of Robert Glaser. Hillsdale, NJ: LEA. pp. 453-494.

Conklin, J. (1987). Hypertest: An introduction and survey. Computer, 17-41.

Davies, D. & Pritchard, A. (1993). The multimedia teleschool&emdash; linking European initiatives. Journal of Computer Assisted Learning, 9, 107-112.

Dempsey, J.V. & Rasmussen, K.L. (1993). Instructional design and development competencies in a new academic program.

Erikkson, G. (1988). Thinking in visual images in the information age&emdash;the changing faces of the school. Gifted Education International, 5 (2), 97-103.

Fabris, M.E. (1993). Using multimedia in the multicultural classroom. Journal of Educational Technology Systems, 21 (2), 163-170.

Forman, D.C. (1995). The use of multimedia technology training for business and industry. Multimedia Monitor, 13 (7), 19-25.

Galloway, Jerry P. (1994) Effect of quicktime multimedia tools on writing style and content. In J. Willis, B. Robin & D.A. Willis, (Eds.). Technology and Teacher Education Annual, 1994. (pp. 94-97). Charlottesville, VA: Association fo the Advancement of Computing in Education (AACE).

Garrett, L.N. & Weiner, B.J. (1998). “Learning on the internet.” Teaching for Success, 10 (7), 4-