Real Learning in Virtual Worlds - CHAPTER 1: Overview
CHAPTER 1: Overview
1.1 Background to the Study
“Imagine waking up in the morning and teaching a class without changing out of your pyjamas. Imagine teleporting and flying to the library instead of inching along a highway. Imagine teaching a classroom of students who may have blue skin, purple wings, or the body of a raccoon. Peculiar as they sound, all of these things are now possible” (Harvard's Berkman Center for Internet and Society, 2007; Kribble, 2007).
With recent advances in public access virtual world technology it is now practical for educators to experiment economically with virtual world based learning methods. Technological limitations no longer impose a substantial compromise on the educator’s preferred teaching method.
Virtual worlds differ fundamentally from the online HTML/PDF based learning environments that have been progressively adopted over the last 12 years for online education in the same way that a book differs from a lecture. At first glance, virtual worlds allow the distance education delivery to move into a virtual representation of the real world lecture, and therefore offer the possibility of a ‘quasi-realistic’ distance education delivery model. At second glance, they tempt the educator who is willing to fund the cost, with visions of highly interactive, immersive and engaging teaching vectors and learning management systems extending beyond the options available in real-world training.
Public access virtual worlds offer educators some potentially significant opportunities in the education space. These include the opportunity to approximate better the real-world education experience for distance learners using low cost (often free) publically available tools, and the reduction in the total-cost-of learning by elimination of travel, reduction in capital (infrastructure) investment through reduction in bricks & mortar infrastructure, world-wide sharing of education content, standardisation of environment navigation and access methods, on demand/automated training session delivery, “24 hours by 365 days a year” availability, instant & automated assessment, instant planet-wide delivery (at homogeneous cost) and the use of software simulations in place of physical models. The virtual reality capabilities of virtual worlds offer immersive exposure to simulations of real-world experiences (like tsunami’s or tornadoes) and events that otherwise could only be described and illustrated in conventional education. They enable exploration of events, places, micro and macro worlds, and theories that are either impossible to do in physical environments, or prohibitively costly to implement for individual courses. Lastly, the use of role play based simulations enable the exploration of foreign locations, cultures and historic events in a manner not otherwise available economically in the physical realm.
As the use of public access online virtual worlds is relatively new to the mainstream education community, many research questions remain unanswered. Exploitation of this technology is still relatively immature compared with traditional online learning platforms and therefore much (although certainly not all) of the content has been more experimental than useful for mainstream educational use – until, possibly, the last few years, if not currently.
Until the last few years, virtual worlds have been either special purpose (like flight simulators) and exceptionally costly to construct, or not sufficiently realistic, difficult to access, complex to use, constrained by limited communication vectors (such as missing audio or streaming media), or cumbersome and expensive to distribute and update. It has only been in the last few years that public access virtual world architectures and infrastructures have reached a level of maturity where convincing workable and low cost solutions have substantially neutralised objections of educators surrounding cost, realism, availability, standardisation, access, content distribution, and richness of sensory and communication vectors.
Possibly the greatest hurdle still faced by educators that are willing to experiment in these worlds is that much of the public continue to perceive public online virtual worlds as game technology. They are yet to be widely acknowledge by mainstream educators as a valid option for the delivery of higher educational course material (Jamison, 2007). Yet the potential for both quality gains and cost savings from the successful exploitation of virtual world training in higher education and industry are very high. There is, therefore, a great need for research in this area that provides insight into the affordances of this technology in education, and guidance on its cost-effective.
While much work has been done to compare the relative “effectiveness” of virtual world versus real world training over many years, little or no structured research has been undertaken comparing the “effectiveness” of different approaches to education within a virtual world.
With a few notable exceptions, research has traditionally examined virtual world training in the context of social interaction or 3D object manipulation and simulation. As discussed in the literature review, this body of work has generally found virtual training to be as effective or better then the real world equivalent (at least within the theoretical confines of the subject matter explored). Yet, a direct consequence of the realism available from the latest generation of virtual world technology has provided the ability to simulate the real world teaching environment itself, not just the ability to build better simulations and 3D models of teachable content. The traditional teaching environment[1] can now be practically reproduced – class rooms or lecture theatres providing a central location for real students to learn in a virtual world. Provided that participants are not constrained by technological requirements as discussed in the literature review, increasing the latest environments allow the reproduction of a real world learning environment, simulating almost verbatim the traditional real world “chalk and talk” lecture experience.
In designing the topic delivery, educators in virtual worlds are now presented with choices between simulating a real world lecture environment delivering essentially similar presentation material to that which they might deliver in a “chalk and talk” lecture in the real world and delivering a purpose built simulation of the material itself – or some combination between these two extremes. The literature review references many studies where the focus has been on assessing the effectiveness of simulation of the teaching material rather than simulation of the real world teaching environment. In the former case the 3D software development effort is significantly in the construction of the topic focussed material, while in the latter the 3D software development is more heavily biased to the teaching environment – such as “lecture rooms”.
Although costs are only superficially explored in this research, it is perhaps reasonable to propose that purpose built, topic centric simulators for each course or subject are necessarily a more expensive investment proposition than a single initial investment in lecture room simulators that are shared by many lecturers and across many topics. The closer the virtual world training delivery model gets to mirroring its real world equivalent the more practical this latter option becomes and the closer the preparation cost matches those of the traditional real world learning methods, yet without the overhead of real world infrastructure and physical student and teacher transportation much reducing the total cost of learning.
A casual survey by the researcher of the teaching infrastructure built by, or for, educators in at least one of these public access virtual worlds that is frequented by more than 200 educational institutions (SimTeach, 2008), reveals that the majority of teaching spaces have been built around exactly this traditional “chalk and talk” lecture model, with essentially conventional auditorium style lecture rooms. Prim-facie this seems an under-utilisation of the environment. Surely, if a 3D representation or simulation of an item can be built, one might argue, the educator is almost duty-bound to exploit the capability. Of course, even in a virtual world with dedicated fast-to-use 3D modelling and agent scripting tools, construction of 3D objects and simulations requires considerably more investment than a simple 2D slide show with audio voice-over, that constitutes the body of a “chalk and talk” lecture.
A central question arises, therefore: on a platform capable of delivering 3D models and simulations, is the mere use of it as a virtual “chalk and talk” class room consisting of 2D lecture slides a reasonable and acceptable use of this technology? This is the central question that this research sets out to explore.
1.2 Research Questions
This study assessed the learning outcomes using two groups in the widely adopted public access virtual world of Second Life. One group experienced a lecture on the topic ‘The Physics of Bridges’ as a 2D slide show presentation and the other group experienced the same lecture as a 3D augmented lecture of the content contained in the slide show presentation. In order to answer the research question: How effective is it to learn in a virtual world using a traditional 2D slide show method compared to that of a 3D interactive simulation?
To carry out this study the following research hypothesis was formed:
Learning outcomes are not independent of the delivery methods in a virtual world, in that varying the delivery method between 2D and a 3D presentation results in a significant difference in the post-quiz achievement scores of a participant in relation to Bloom’s cognitive process of factual knowledge of ‘remember’ and ‘understand’.
Second Life was chosen as the experimental platform for the research question because of its low cost of access (free), wide platform availability (PC/Linux/Mac), it wide adoption (16 million plus registered users (Linden Lab, 2008a)), huge educator community (200 plus educational institutions (SimTeach, 2008)), maturity and capability of its tool set (3D, streaming and interactive audio, streaming media, web interfacing, html content support, etc), content publication delay (instant) and its environmental realism (real time content streaming, spatial audio, environmental and spatial lighting, 3D perspective, layering, animation, concurrent multitasking agents, realistic photo finished avatar mesh, etc.).
1.3 Overview of Study
This research study was conducted in the online virtual world of Second Life. Using an experimental design approach a virtual learning campus was constructed to utilise two different forms of lecture delivery method on the topic of ‘The Physics of Bridges’. This topic was presented as a lecture with a 2D slide show and audio (reproducing a real world lecture on the topic in the virtual space) and the same 2D content and audio augmented with immersive 3D models. Both delivery methods used identical content, slides, audio and time allocation. The independent variable in the delivery method was the presence or absence of 3D bridges and simulations matching the 2D slides and audio.
The 2D and 3D lecture environment simulated real-world lecture theatres with seating for up to 18 people and a large front facing projection screen. The 3D lecture room contained an additional space with lecture screens on three walls in which 3D objects appeared and with which users could interact or examine. The 2D and 3D theatres were otherwise identical.
Participants were recruited from the in world population of Second Life by advertisement and self selection (i.e. without profiling or filtering) and without replacement (avatars could not repeat any test). Prior to the lecture they received a pre-quiz containing 8 questions to establish a prior-knowledge benchmark. After completion of the pre-quiz participants were randomly allocated to either a 2D or 3D lecture theatre. On completion of their lecture they were given a 20 question post-quiz to test the learning outcomes of the lecture and a survey to gain an understanding of their learning experience within the virtual world environment. A total of 111 participants took part in this entire research process. The 2D and 3D participants numbered 55 and 56 participants respectively.
The learning materials along with the pre and post quiz questions were constructed using Bloom’s cognitive processes of ‘remember’ and ‘understand’. The quiz questions were divided evenly across these processes, which provided the basis for analysis.
The analysis method adopted in this research was triangulation using mixed methods. The pre and post quiz questions provided the basis for quantitative analysis. The post survey open questions provided the basis for qualitative analysis. Both of these analyses were then triangulated in order to compare the learning outcomes and experiences of the two groups that took part in this research.
1.4 Significance and Limitations
Mirroring real world education, there are at least three barriers an educator must overcome in order to deliver virtual world training:
Hosting infrastructure (the software environment the hosts the virtual world mechanics) Training infrastructure (the creation of training spaces in the virtual world such as lecture theatres, or presentation screens) Training content (the actual training material presented).
Today’s public online virtual worlds provide the hosting infrastructure while enabling low cost construction or acquisition of the training infrastructure to enable educators the opportunity to experiment with virtual learning delivery efficiently. Prior to this, educators were faced with extensive time, cost and complexity to build custom applications that could deliver the infrastructure before any virtual learning could take place. What once required extensive support from heads of department now requires very little effort on the educator’s behalf to enter into the world of virtual learning.
With a public online virtual world such as Second Life, the cost to develop, publish and deliver a 2D slide show based instructional learning program, such as the one produced in this experiment, is comparable to that of a real world ‘face to face’ lecture. Yet given the opportunity of this technology to go beyond real world instructional methods the temptation to exploit the full modelling and simulation capabilities of the environment is strong.
The research aimed to inform the question as to whether it is ‘worth’ the extra cost and time to build something more complicated than a 2D slide presentation. For this research the cost was measured in time (hours). While the cost of the 2D lecture was identical to preparing and delivering the same in the real-world, the 3D augmented version was approximately 3 times the cost. There is therefore a significant incentive to determine under experimental conditions the difference in learning outcomes and experience of the participants when presented with two different forms of delivery methods.
To preserve the integrity of the concept of separation of costs of content from costs of infrastructure (both hosting and training), a re-usable general purpose campus and lecture space was first constructed. With all content and tests independent of the campus and lecture infrastructure and interchangeable, the environment that can support both multiple simultaneous courses and rapid 5 to 15 minute course change in each lecture room. While this was not critical to the study, it was judged essential to the integrity of the assumptions on which the research was based: that virtual world content could be treated independently of the training infrastructure if a shared protocol was adopted. Secondly, the content preparation technology expectations were intentional constrained to a standard SL and MS Office equipped PC. PowerPoint and MS Audio Recorder (or other audio recorder) and the Second Life client are all that is required at the minimum to prepare a course for delivery for the purpose of the research.
Despite the recent growth in publicly accessible on-line virtual worlds, little published work has been conducted in this specific area of research. Furthermore, at the time of writing none, if any, had been performed using experimental methods. There is a growing body of high grade and scientific work in other aspects of educational and social dimensions of virtual worlds, and a respectable body of earlier work in purpose built and text based 3D virtual worlds, particularly in the comparative aspects of virtual and real world presence. Possibly, it is only with the realism attained in the latest generation of full content streaming, mixed graphical, audio and text worlds that this research has become practical. Thus the researcher’s motivation is to add to a body of knowledge, which is, as yet, predominantly (if not totally) lacking in scientific rigour via an experiment conducted under controlled conditions.
There have been multiple studies that compare traditional face to face learning methods with distance education learning outcomes. Thomas Russell’s book ‘No Significant Difference Phenomenon’ (2001) documents a review of literature of accumulative studies that goes back as far 1928 with the research question: ‘Does taking a course via distance education lower a student's chances for success as compared to the same student taking the same course in a face-to-face format?’ In most cases Russell’s findings resulted in ‘no significant difference’ in learning outcomes. The common identifier by Russell being that no student is better or worse off when comparing distance learning delivery methods with that of traditional face to face learning methods.
Similarly Richard Clark’s (1983) article published in the early 80s ‘Reconsidering Research on Learning from Media’ claimed that when comparing learning effects of different media platforms, there is no signification difference in outcome. In this article, Clark dismissed any studies that did find differences by providing that any differences that may have been found were not due to the medium platform but rather to the instructional design in the study.
Clark’s article sparked a heated response from Robert Kozma who had opposing views on the matter. This lead to a public debate between the two researchers (R. E. Clark, 1994; Kozma, 1994) in academic journals. This debate continues today amongst educational researchers and is commonly termed ‘The Media Debate’ (EduTech Wiki (2009).
This researcher does not enter into the media debate nor does she enter into the debate over whether real face-to-face learning ‘is better’ or ‘worse’ than virtual world learning. Rather this research has taken the position of ‘Now we are here [in the virtual world] what do we do?’
Consistent with this position the research decided to recruit only from the in world population. Therefore the constraint related to this is that the tested population is more likely to be pre-disposed to the virtual environment for a range of purposes one of which might include education. In the context of this experiment, however, the researcher is not convinced that such a condition would have had any impact on the outcomes. The elimination of the novice user dimension removed mechanical unfamiliarity as a significant factor from the outcomes which was appropriate for a study comparing virtual world delivery methods as opposed to a study comparing virtual and real world learning methods, and has been a factor that complicated the interpretation of some virtual-world research results in prior studies.
1.5 Structure of Thesis
For common terms used in this thesis see Appendix A: Terminology.
Chapter Two Literature Review; examines virtual world technology and a brief overview of educational learning theory.
The Virtual world section discusses alternative definitions, characteristics, history, key architectural features, research outcomes and applications in education of virtual worlds. The review of virtual worlds has been taken from an historic perspective discussing key influences that have lead to today’s massively multi-user virtual worlds. Discussion of virtual worlds concludes with a review of educational uses, affordances and a review of current research into online virtual worlds.
Chapter two concludes with a review of learning theory and instructional methods that provides basis of the learning methods and materials used to conduct this experiment.
Chapter Three Research Design; examines the research design along with the researcher’s theoretical assumptions, environment design, lecture material design and analysis methods adopted in this research study.
Chapter Four Results: presents the quantitative and the qualitative results of the virtual world learning experiment conducted in Second Life between the two groups of participants who undertook the differing lecture delivery methods for a lecture on ‘The Physics of Bridges’.
Chapter Five Discussion & Conclusion; provides an analysis of the results of the experiment along with discussion of these results and opportunities for further research.
