Regarding the abductive reasoning component, an example from Carbon World is included in Table 17, below.

It also became clear that students constructed culturally mediated stories; that they found a room for their individual contributions within the development of a communal voice, which gave rise to what was in essence a visual, interactive language that has meaning, particularly to the group who created each individual environment. To be respected as an individual, to be heard, is as Coles (1989) describes, one of the most elemental aspects in the development of self worth. He states:

In fact, each individual data set for each student tells a slightly different story. It was clear from the results that different students reacted differently to the project, that they had indeed constructed an understanding of their own, that was illustrated on the pages and video tapes that we collected. But most importantly, that understanding resided inside each individual and the collective memory of the group.

5.1.3. Treatment Discussion

In this section, interpretation of treatment results from concept map analysis are provided.

5.1.3.1. Interpretation of Constructivist vs. Traditional Treatment Findings

The reasons for the lack of significant distinction between the Constructivist and Traditional treatments could be attributed to the following:

1. These students were already constructivist learners, even when taught traditionally.
2. There was so much to learn under the constructivist treatment that the students may have been experiencing cognitive overload.
3. The potential for superior learning in a virtual environment is related to a high degree of presence, which was lacking along some dimensions for most students.
4. Exposure to the virtual environment was brief (< 5 minutes), and navigation and interaction was difficult for some students.
5. Cognitively able students learn regardless of pedagogical considerations.

Regarding the students’ daily classroom activities, these students took part in thematic, cross-subject, project-based learning as described by Zemelman, Daniels & Hyde, (1993) and Brooks & Brooks (1993). In fact, the case can be made that these children were already richly steeped in both collaborative and individual learning opportunities. They were living the kind of inquiry-based learning practices espoused by both Cunningham (1992) and Shank (1992, 1997).

In asking these students to learn two of the four cycles in the ‘traditional’ classroom, using traditional teaching tools (lecture, textbooks, and worksheets), it is possible that the students took their day-to-day learning practices from the constructivist classroom into the traditional classroom with them.

In fact, the only additional aspect of constructivist learning that this project brought to these students was the opportunity to be the designers of a knowledge base from the ground up; to learn how to use the technology, to model in 3-D, to consider how to design and develop an interactive educational environment, and to embody textual information in a visual format. However, this was a formidable task. It could have been that there was too much world-building activity, negating the students’ ability to absorb more content information during their constructivist treatment sessions.

This relates to the second point in the list above, that students may have been experiencing cognitive overload. The most common complaint about the project was that students felt they did not have enough time to adequately design, develop or experience their environments. Data should have also been collected on whether students felt they had enough time in the traditional classroom. Survey results indicate that students disliked the research component of the project the most. In the traditional classroom, they were almost spoon-fed the answers. For example, the worksheets had page numbers for reference on them. In comparison, students had to look up everything of relevance themselves in the constructivist classroom, in addition to all of the other skills that they were learning.

It remains to be seen what students can do when they do have enough time. We have yet to be in a classroom environment where we were not in a rush to get everything done within a very tight time schedule. Others studying the effects of virtual learning environments (Dede, et al, 1996; Loftin, et al, 1993) have had more leeway over their development schedules, leading to a different set of issues that they have been able to more adequately address.

Survey results regarding the experiential portion of the project indicate wide differences amongst students regarding their sense of presence. Subsequent research indicates that superior learning in a virtual environment is tightly coupled to a high degree of presence and that when the sense of presence is reduced, so is the opportunity for learning. (Winn, 1995).

Students’ sense of presence in this study may have been limited due to lack of experience in the virtual environment. They had no opportunity for practice in a virtual environment, so that the first time they went into their virtual space, they also had to learn how to navigate and interact with virtual objects. This is not a good way to introduce students to a virtual learning environment (Moshell, 1995). As is true with computer-based interface, if you are too busy playing with the buttons, you can’t really enjoy the show (Kay, 1990; Laurel, 1990).

Regarding high-ability vs. low-ability students and the effects of virtual environment construction and experience, Winn (1997) has found that low-ability students, particularly male low-ability students, benefit most from the kind of constructivist approach followed in this study, and that high-ability students tend to learn regardless of the teaching style or classroom environment. Low-ability students often require either additional time or assistance to complete their tasks when designing virtual learning environments (Winn, 1997). However, in this project almost all of our students were either average or above-average ability, and most were also highly self-motivating, as described in the KCOT program application. (Kellogg Middle School, 1996). We had only three low-ability students, and those that we did have were all in the Carbon group. The pre- and post-test results of the Carbon group were consistent with Winn’s (1997) findings for the world-building portion of the project.

In interpreting the concept map data more broadly, it should be mentioned that the process of performing visio-spatial exercises, such as drawing, modeling, and visualizing objects in three dimensions coupled with actually modeling objects on the computer had a profound affect on students’ concept map representations and on the depth of understanding associated with the wetlands processes they represented. As described by Samuels & Samuels (1975), Morris & Hampson (1983), and Adams (1989), imagery is an essential tool used to understand visual and spatial relationships. Furthermore, the process of translating information between symbol systems (Mones-Hattal & Mandes, 1996; Adams & Hamm, 1988) results in the utilization and enhancement of one’s higher level thinking skills, as described by Bloom et al. (1956).

As stated by Cunningham (1992; 1997), signs can be highly independent of what they reference. As is true of developing any tightly-woven referential system, the more signs that link to objects of interest, the better recall an individual has. If this study has provided these students with the opportunity to utilize their visualization skills more fully, by providing alternate access to information, then it has been successful. The increased use of pictorial and diagramatic concept map representations may well be a sign of this success.

Students had least difficulty developing representations about concrete, physical relationships, even though the icons and indexes used to represent these relationships were somewhat abstract (lightening for energy, fox for all carnivores, etc.) The students definitely had a more difficult time considering chemical relationships, and how to represent them. Discussion about the representation of oxygen or nitrogen took at least twice as long as did discussions about the representation of water vapor or algae.

In teaching young people physics, Minstrell (1992) finds students have a strong attachment to specific (physical) objects and to less developed modes of reasoning during their younger (pre-teen) years. As they grow older, their ability to develop abstract representations becomes more pronounced; findings similar to those presented by Perkins (1993). It is this ability to abstract that adds rich new dimensions to the meaning-making process (Cunningham, 1992).

Finally, there was insufficient time to properly provide students with deep lessons and practice in abductive reasoning (Cunningham, 1988; Shank, 1989), which may also have led to the lack of differentiation between the constructivist and traditional treatments. The original intention had been to take the time for iterative design, and to provide students with an opportunity to teach each other and to teach non-KCOT students using their virtual environments. Both of these activities would have led to greater opportunity to develop hypotheses and to test one’s assumptions about most aspects of the virtual learning environment. Unfortunately, the schedule did not permit this kind of deep, intensive inquiry.

5.1.3.2. Interpretation of Constructivist vs. No Instruction Treatment Findings

Significant differences were found between the constructivist and the no instruction treatment results, as expected. Even with the overlap between cycles, it was expected that students receiving no instruction in a given cycle would certainly do less well than those receiving instruction of any kind.

Of course, this was not the case in comparing the traditional vs. no instruction treatments, as described below.

5.1.3.3. Interpretation of Traditional vs. No Instruction Treatment Findings

Another unexpected finding was the lack of significant difference between the traditional and the no instruction treatments. However, the concept map data, which provided me with the bulk of my treatment analysis information, was skewed by the preponderance of students drawing the water cycle. Students knew this cycle whether they were studying it or not, and it was definitely the cycle of choice for the ‘chosen’ rather than ‘built’ representation. Some students also knew the energy cycle, which enjoyed second-place billing for the number of ‘chosen’ representations.

There is no basis for a strong treatment-based comparison between the traditional and no instruction options. Had the depth of the students a priori knowledge of the water and energy cycles been known, compensating factors would have been put in place, such as constraining the represented (chosen) cycle to tie to one of their ‘traditional’ treatment cycles, or to have had students draw concept maps representing the cycle associated with all three treatments.

5.1.2.4. Other Observations

What the quantitative data do not describe is how very much the students’ language and presentation techniques changed and grew over the course of the project. Students began to speak in terms of their ‘perspective’, and ‘rotating their view’. They seemed more willing to consider part-to-whole relationships in their other classes. All four KCOT teachers noticed this trend.

The concept maps, as discussed above, show a clear movement towards the incorporation of visual metaphors in their post-tests, which can be attributed dirctly to their virtual environment construction process, as this was the only component of the project that included visual representations. It affected the manner in which students chose to represent information regardless of what instructional paradigm had provided the initial content.

Alternative assessment provided a means to get at the heart of what became meaningful constructions for both individuals and for each group. This became clear when working with the special needs students present in the KCOT classroom. For example, one very shy, intellectually challenged 12 year old girl managed to create a fox for Nitrogen World. This was her first "creation" or "performance" of the year; the first indication that she was able to construct understanding about a concept in a way that allowed her to contribute both to her personal knowledge base and to a larger, more collaborative construction.

This was a phenomenal accomplishment for her. However, when compared to the larger classroom of students, who were all contributing objects, providing constructive comments, and determining interrelationship possibilities, this one student didn’t really measure up from a "performance" standpoint. However, in utilizing a variety of measures and by valuing the process of self review (Reif, 1990), students maintained their sense of self-esteem and motivation throughout the study.

5.1.3. Opportunities for Further Research

5.1.3.1. Overview

Findings indicate that there was no significant difference between the Constructivist and Traditional treatments. Though significant differences were not found between the Constructivist and Traditional treatments, nor between the Traditional and No Instruction treatments, subsequent research has been conducted that supports the original hypotheses put forth in this study (Winn, 1995, 1996, 1997; Taylor, 1997, Osberg, 1997; Dede, 1997). Additional research opportunities based findings discovered after completion of the VRRV project in toto have been included in a special addendum. Suggestions for additional research provided in this section of this chapter are based on findings specific to this study.

Teaching by traditional means is a well understood endeavor, as it has been the norm for hundreds of years. However, there is opportunity to further test the educational value of constructivist practices. If at minimum constructivist practices do no harm, they are certainly worth researching further. We should build on our understanding of what aspects of constructivist practice may provide additional value for the student. Therefore, this study provided a good starting point from which further research can be conducted into constructivist practices in the classroom, especially those utilizing virtual technologies as an adjunct learning tool.

It is clear from the results that students made meaning from their knowledge constructions under both the Constructivist and Traditional treatments. This is a semiotic issue. Students created and used icons, indexed and symbols extensively, and engaged in abductive reasoning, as presented in Tables 16 and 17, above.

Even though it is clear that students created rich visual and interactive representations, and reasoned abductively in their virtual environments, semiotic issues could have been more deeply addressed. Further research needs to be completed into the nature of the knowledge construction process, as described by Cunningham (1992), Shank (1989) Phillips (1995). It would be valuable to relate that process to the creation of both tangible and virtual objects, how they are developed, and how they come to have meaning, both as symbols and as directly accessible objects (Mones-Hattal & Mandes, 1996). In addition, a rubric could be designed to assess the richness of particular signs, and of sign systems. This could lead to the development of a more universal visual and interactive semantic that would have the potential to transcend cultural boundaries, while still allowing for individual creativity in the design and development of individual signs.

From a Constructivist perspective, additional research could be conducted on the general value of constructivist learning; learning for depth vs. breadth (Brooks & Brooks, 1993), how and when to incorporate visual and verbal representations (Mones-Hattal & Mandes, 1996), whole body learning and experiential education (Hutchins, Hollan, and Norman, 1986; Kraft & Sakofs, 1989), meaningful self-directed learning (Poplin, 1991), and examining one’s reasoning for developing certain knowledge constructions (Minstrell, 1989; 1992; Minstrell, Stimpson & Hunt, 1992).

Metacognitive Issues

One goal of constructivism is to teach students how to effectively question the information placed before them. Pressley, Harris & Marks (1992) discuss the development of metacognitive strategies, as couched within a constructivist framework. Though there are those who believe that critical thinking skills can only be taught within the confines of a content domain, others feel that one can learn basic strategies that can be applied across content areas (de Bono, 1991; Salomon, Perkins & Globerson, 1991; Scardamalia, et al., 1989; Butterfield, Wambold & Belmont, 1973).

Process Issues

Another opportunity for additional research has to do with existing classroom practices. At Kellogg Middle School, the VRRV team worked directly with the four-classroom program that was already utilizing a constructivist approach to the learning process. An intensive three to four week program of this nature undertaken in a more traditional classroom environment might yield more substantial differences when comparing Constructivist and Traditional treatments. Since the Kellogg students were already used to learning in a constructivist fashion, it is quite likely they just kept on utilizing the perspective and the cognitive tools that they used in their regular classroom environment, regardless of treatment. If one were to conduct further research on the educational effectiveness of constructivism, it seems clear that the comparative value would be enhanced by starting from a more traditional position.

Understanding Virtual Reality as a Learning Tool

From a perceptual perspective, additional research could be conducted within the virtual environment itself, testing navigational paradigms, effective use of color and texture, spatial manipulations of scale and orientation, and the more prevalent and effective use of auditory and haptic feedback. Furthermore, research into developing meaningful virtual tools would be useful (Rose, 1996), as well as designing and testing new navigational paradigms. All of these opportunities involve the use of signs and metaphors to make meaning in a virtual space.

On the practical side, the access and administrative aspects alone require in-depth analysis if the world-building process and virtual reality technology are to ever become an integral part of the learning process.

5.1.4 Other Issues

5.1.4.1. Adding Depth to the Existing Study

There were also many areas where, with more time and human resources, additional depth could have been added. Some examples include:

• Providing students with an opportunity to teach others using their wetland environments, to see what would be meaningful when trying to share their environments with others.
• Providing extra time so that initial designs could have been prototyped, reviewed, and modified rather than having to go with the ‘first cut’ due to time constraints
• Using stronger assessment tools, that would have provided a more fine-grained analysis, particularly between treatment groups.
• Providing more opportunity for cross-group discussions, to help the students develop more robust environments
• Providing tools and conducting analyses that would have yielded a deeper understanding of the meaning-making process and its affective components, as opposed to the knowledge-construction process upon which we focused our efforts.

There is not a substantial enough body of research to ascertain exactly what makes a virtual environment useful and enjoyable, when to use virtual technologies rather than other media, and the circumstances under which virtual environments are better learning tools than any other way to come at the learning process. Additional understanding could be garnered in these areas by conducting research in the areas listed under the bullet points above.

5.1.4.2. ‘Covert’ Learning

Survey data indicated that the educational component of the process for some students was not as much to their liking as the creative and technical aspects. This sets the stage for a discussion regarding ‘covert’ learning. The teachers, and the HIT lab staff were diligent in their presentation of a balanced perspective regarding the value of both the educational and technical components of the project. For 20% of the respondents, the educational component was rated poorly. However, it is clear from the test scores that the subjects learned about wetlands, regardless of whether the educational component of the project was presented using traditional means, or learned using constructivist means. However, since all of the subjects experienced the constructivist/world building treatment, and most of the resulting concept map illustrations indicated that this treatment had a deep effect on how they represented information, it can be assumed that the constructivist/world building activities directly affected their learning process.

This raises the question of using the world building and experiencing component of the project as a motivator for covert learning. Specifically, using virtual environment development may be a useful tool for those students who perhaps do not like, or do not respond well to traditional classroom practices. Subsequent research has indicated that this is indeed the case; that world building and experiencing is especially motivating for lower functioning boys (Winn, 1995), and is also motivating, though less so, for higher functioning students. At Kellogg Middle School, this was shown to be true as well, with regard to the Carbon group, in which the three lower functioning students were placed.

Regardless, the Survey indicates that almost all students enjoyed most aspects of the project; that the positive aspects far outweighed the negative components of the project for most of the students. We often heard students say that they felt that the project was moving too fast; that there wasn’t enough time for them to accomplish what they wanted to accomplish. Even though only 8 individuals mentioned lack of time specifically on the survey, it was the complaint most often openly voiced; not enough time to learn all about the modeling software, not enough computer time to create all of the objects the children wanted to develop, not enough time ‘under the helmet’, or not enough time for the project as a whole.

5.2. Conclusions

In conclusion, results indicate that the world-building process, coupled with the opportunity to experience one’s virtual learning environment is a powerful, motivating way in which to learn about wetlands ecology. Apparently the traditional educational environment provides an equally educational experience, given the way the research program was designed.

On the Constructivist side, findings indicate that by incorporating the student’s creativity and design skills, metacognitive skills, freedom to make their own design and navigational decisions, and their whole body into the learning process, students have a very wide avenue of opportunity for cognitive, somatic and affective growth and experience. On the Traditional side, teacher-led lectures, textbooks, and worksheets appear to be equally educationally valuable.

The field on both sides of the fence is rife with opportunity for both research and development. We must test and re-test our assumptions and how they affect our developments, and to encourage and support student involvement from idea generation to usability assessment.

To this end, a two-pronged approach to additional research is presented that incororates traditional educational programs for the students who do not want to take part in a constructivist environment, and constructivist classrooms that incorporate virtual educational programs that allow students the opportunity to participate in and contribute to the development of a virtual educational network. As one female student from this project said: "I don’t think we are ready for this technology; our teachers don’t know enough about it yet." Shank (1992) would agree. But the students do. Interestingly, this is the method used by Kellogg Middle School. They provide 6 different tracks or programs that vary in type of instruction, use of technology, core course concentrations, and time on task.

Several students mentioned the difference between verbal presentation, and creating visual, interactive cause and effect relationships themselves. One male student said "I didn’t understand the nitrogen cycle when the teacher explained it. I do now!!". A female student said "It’s harder to learn this stuff out of a book; here I could go with it as it was happening. I was in control!" Another male student said "I knew absolutely nothing about this prior to building my world in VR. I used it to learn. I learned a process. It was fun, so I’ll remember it more.", and another said "I understand it better now that I’ve experienced it."

However, 20% of the students did not like the research component of the project. It would have been good to have conducted some visio-spatial testing prior to starting the project. It has been found (Winn, 1997) that high spatial students enjoy and value their virtual experiences highly. It would have been interesting to see if high spatial students correlated with those finding the research process onerous. If so, perhaps using the technology as a motivator is a completely valid approach to balancing a students verbal and visio-spatial development.

All of the students wanted more of everything; more time, more realism, more water, more mud, more animals and plants, more behaviors, more sound, more applications, more environments. I for one intend to try and give it to them.