Educational Technology Research and Development

, Volume 61, Issue 6, pp 979–999

SciEthics Interactive: science and ethics learning in a virtual environment


    • School of EducationIowa State University
  • Joan Woolfrey
    • Department of PhilosophyWest Chester University of Pennsylvania
  • Matthew Pierlott
    • Department of PhilosophyWest Chester University of Pennsylvania
  • Seth Kahn
    • Department of EnglishWest Chester University of Pennsylvania
Development Article

DOI: 10.1007/s11423-013-9319-0

Cite this article as:
Nadolny, L., Woolfrey, J., Pierlott, M. et al. Education Tech Research Dev (2013) 61: 979. doi:10.1007/s11423-013-9319-0


Learning in immersive 3D environments allows students to collaborate, build, and interact with difficult course concepts. This case study examines the design and development of the TransGen Island within the SciEthics Interactive project, a National Science Foundation-funded, 3D virtual world emphasizing learning science content in the context of ethical dilemmas. The 2 year development process is examined through the lens of the rapid prototyping instructional design model, following the project from conceptualization to implementation of a 3D simulation. Through expert interviews, focus groups, and working groups, we were able to determine critical scientific and ethical issues to present to learners in the virtual world. We collected data on 53 students using the simulation at universities in the United States and South Africa and evaluated their experience using qualitative and quantitative methods. Results showed that student participants were engaged and motivated by the simulation. The students reported an increase in science knowledge and ethical understanding, but individual experiences varied.


Virtual environmentScienceEthicsRapid prototypingInstructional designOpensim


Increased computer and internet access in higher education over the last few years has greatly expanded the choices available for faculty wishing to explore new and emerging technologies for teaching and learning. While online learning software, such as Blackboard and Moodle, enable faculty to use multimedia tools and text to enhance learning activities, 3D virtual environments take online interactions to another level; students can walk, talk and create as a community of learners. An increasing number of educators are using these spaces in K-12 and higher education (Hew and Cheung 2010; Warburton 2009; Wimpenny et al. 2012). Students can create, collaborate, and communicate from around the world at any time of the day.

This article reports on a National Science Foundation (NSF)-funded virtual world project. SciEthics Interactive is a training simulation to teach science content in advanced undergraduate science courses while simultaneously increasing awareness about ethical issues that arise regularly in scientific practice. The goal of the project was to address on-going concerns that new researchers are not being taught to see interconnections between moral responsibility and the conduct of research (Hollander 2009; Mayer and Steneck 2012; Nichols-Casebolt 2012; Rappert 2010; Smith-Doerr 2006). Using the virtual world for training in ethics and educating in science at the advanced undergraduate level is a novel use of the virtual platform (Houser et al. 2011; Hai-Jew 2011). Our project suggests that the virtual world is an ideal platform for such explorations since, among other features, it avoids real-world consequences from less-than-optimum decision making. In cases where poor judgment could lead to real harms, having a virtual world in which to practice decision making with ethical implications can be especially valuable as a training ground.

This article reports on the design, development, and implementation of TransGen Island. In what follows we will provide a description of our process in hopes that our experience might prove useful to others embarking on similar projects. We also discuss the project context, including the concerns that have motivated the NSF to set aside funds for such projects in the first place. In detailing our development and design process, we use the rapid prototyping instructional design model. In conclusion, we discuss the initial results of our project and offer recommendations and “lessons learned” for others who might be interested in developing similar projects.

Background and context

Science and ethics in virtual worlds

The players in virtual environments, called avatars, participate in communities based on common interests. Students are able to interact with objects too big, too small, too far away, too expensive, or too dangerous for the classroom. The ability to create complex interactions allows for authentic student experiences in a safe environment. For example, K-12 and higher education institutions have used virtual worlds to explore ecology (Hickey et al. 2009; Wrzesien and Alcañiz Raya 2010), chemistry (Lang and Bradley 2009), and medical sciences (Beard et al. 2009). Other activities include a jet propulsion laboratory, a giant cellular structure, tsunami simulation, and a virtual hallucination for medical professionals (Second Life 2012).

Ethical training is an important cornerstone of science and engineering education (CSEPP 2009), but “the use of virtual worlds in teaching ethics in higher education has not been reported” (Houser et al. 2011). Although there are some examples of virtual worlds with content related to ethics in advertising (Alperstein 2011), theoretical foundations of ethics (Houser et al. 2011), ethics of virtual worlds (McArthur 2008), and ethics in middle school science (Barab et al. 2010), SciEthics Interactive is the only classroom activity in the research literature designed for the exploration of authentic ethical issues in the context of university-level science and engineering.

Ethics education in the sciences

In the non-virtual world, there is extensive literature on issues related to ethics training in the sciences (Mumford et al. 2008; Speight and Foote 2011; Steneck 2004, 2009, 2013). In general, these studies emphasize research ethics. The ethical concerns that get the most attention in research ethics literature—those the federal government deems “serious research misconduct”—involve falsification, fabrication and plagiarism in the generation of scientific research (CSEPP 2009; Helton-Fauth et al. 2003). Focus on ethics in the sciences also frequently centers on issues of integrity: of both the researcher and the research data. Integrity, as it is often defined, emphasizes values deemed important for all researchers, such as honesty, accuracy, efficiency and objectivity (Steneck 2004, 2013). Scientists call upon each other to foster responsible conduct among peers and subordinates, including being competent in one’s field, maintaining fairness, ensuring the appropriateness of all one’s relationships (e.g., researcher/participant or professor/student, etc.), and respecting the individual in human subjects research through informed consent and confidentiality (Woody 2008).

Despite efforts within the scientific community and despite 25 years of federally-mandated attention, reports of serious research misconduct are on the rise, and the pipeline for researchers is filled with “millennial” students who are increasingly blasé about plagiarism and other “academically dishonest activities” (McCabe 2005; Titus et al. 2008; Titus 2010; Wright et al. 2008). The NSF, concerned with the increase in reported instances of research misconduct, sets aside a percentage of funding for projects that integrate ethics into science curriculum. The NSF’s Ethics Education in Science and Engineering (EESE) program found our project to be innovative and intriguing, therefore funded our proposal in 2009. This project utilizes a 3D computer environment to engage and motivate students in authentic learning activities in order to address these challenges.

Consistent with the conclusions reached by moral psychologist James Rest (1979), Helton-Fauth et al. (2003) identified four components necessary for ethical decision making in any context: (1) problem-seeing, (2) formulating judgments, (3) motivation, and (4) ethical action. The work of these psychologists also highlights the reality that, not surprisingly, competency in a field correlates with an awareness of ethical concerns in that field (Woody 2008). In other words, understanding of what a discipline requires will entail understanding the importance of accuracy in reporting data. You will then be committed to efficiency in your work and objective analysis, and honesty will motivate your conduct.

One important lesson learned from Helton-Fauth’s work is the more ethically ambiguous a context, the more genuine the responses from the participants. If the problems the individual must contend with are ill-defined and solutions are far from clear, responses to the problems will more closely approximate real-world responses (Helton-Fauth et al. 2003). This is one place where virtual worlds can play a valuable role for ethics training. One of the guiding principles for the development of the ethics portion of this project was to raise ethical dilemmas to which there were no clear answers because ambiguous situations generate the most “real life” responses. In addition, asking the participants open-ended questions about the ethical implications raised by their virtual world activity exposed participants to ethical issues they may not have considered before. Since “problem-seeing” is a fundamental step in both Helton-Fauth’s and James Rest’s work, just by asking the questions, we are creating the space within which moral awareness can be triggered. When used in a pedagogical context, the questions asked in the training simulations prompt instructors and participants to dialogue, comparing experiences and examining assumptions about the values and principles at play. The manuals developed by the team, an ethics primer for the instructor and handbooks provided to instructors and students, help shape and guide that conversation. If moral integrity is defined by the capacity to do the right thing, such that they would act rightly out of desire as well as awareness, then the four components mentioned above are useful for understanding how to encourage such integrity in any context, including the science laboratory or classroom.

Additionally, there is compelling evidence that actively engaging students in learning and applying concepts in multiple real-world contexts greatly improves student learning outcomes in science (DeHann 2005). Innovative teaching approaches also improve the confidence and performance of under-represented groups in science (Weston et al. 2006). Thus, virtual world simulations that help students learn science content interactively, in the context of ethically ambiguous (i.e., realistic) situations, are likely to increase student understanding of science content while increasing their appreciation of the wider relevance of abstract scientific and ethical ideas.

SciEthics Interactive aims to address needs in science education for science content knowledge as well as ethics training, with the goal of seeing the two more thoroughly integrated. The SciEthics Interactive project has generated (i.e. designed, built, and tested) virtual world simulations which can be incorporated into science courses and research mentoring, for the purpose of modeling and developing ethical behavior that contributes to the “who we are” of the next generation of scientific researchers.

SciEthics Interactive

The SciEthics Interactive project was developed by a team of researchers with expertise in the fields of ethics, educational technology, science education, and qualitative classroom research/research ethics. The areas of expertise complemented the various stages of project development: conceptualizing, data-collecting, designing, testing, and revising both the simulations and assessment of implementation. There are four different constituencies in this project: local science faculty (Faculty 1), the grant team (the team or project team), student programmers (student assistants), and science faculty from various institutions around the country and from South Africa (Faculty 2) who piloted our virtual simulations. Because the team is comprised largely of non-scientists, they partnered with Faculty 1 who specialized in various scientific fields to develop sound science content. The content knowledge and experience of the science faculty were tapped in individual interviews and hands-on working group sessions, the results of which the team transformed into a manageable vision that was then re-imagined within the virtual world by the educational technology faculty member and a team of student assistants.

The SciEthics Interactive activities are aimed at advanced college-level science or engineering courses. Instructors provide students with a detailed information packet, including login procedures, background information, a task list, and assignment instructions. The activities require very little knowledge of the virtual world by the instructor, alleviating the steep learning curve typical of teaching in virtual environments. The instructor also has access to a faculty handbook with specific instructions for preparing students for the assignment. An “ethics primer” discusses the interconnection of science and ethics. The handbook offers context, theory and vocabulary for leading a classroom discussion after the students have performed their task. Instructors assign the virtual activities as a supplement to their course curriculum. These activities are expected to take students 2–5 h of out-of-class time. Since the simulations are web-based, they are especially flexible and adaptable to a variety of courses and curricula. The flexibility extends to the collaborative element of the activity, since students can work as individuals or in groups without changing the learning goals.

Since receipt of the NSF grant in 2009, the research team has developed and released the virtual world simulation TransGen Island (Fig. 1). TransGen Island is designed as an individual activity with three role-playing identities (i.e. research scientist, activist, or government regulations agent) to use during the simulation. Although the roles are different, each student is asked to perform similar tasks throughout the activity. Focused on the issue of the genetic modification of salmon by the company TransGen, students arrive on the island and enter the Orientation Area where they become acquainted with their avatar, learn how to navigate in the virtual world, and acquire knowledge of the layout of the island. Next, the students move to the GMO Museum where they review slides containing information about the history and science behind genetic modification of plants and animals. In the main building, students visit with a non-playing character (virtual) receptionist who guides each student to supplies and offices appropriate for their assigned role. Students complete the virtual portion of the activity by collecting measurement data in the labs, exploring other areas of the island, and taking photographs. Following these science-related activities, each student must walk down a path containing ethical questions, such as “How concerned would you be if these fish were to escape into the wild and mate with genetically pure salmon?”. They then take a survey on perceptions of the assignment and learning before leaving the virtual world. The activity ends with the submission of a final report to the course instructor, written from the perspective of their assigned role on this island. Students are asked to provide evidence in this report for their recommendation as to whether the TransGen salmon product should be approved for human consumption.
Fig. 1

Map of TransGen Island

Instructional design process

Rapid prototyping instructional design model

The design of a virtual learning environment must not only take into account the best practices in instructional design, but it also must pay very close attention to the technical considerations. The rapid prototyping instructional design model (Tripp and Bichelmeyer 1990, Fig. 2) successfully blends the two concepts in a flexible approach. Originating from software design methodologies, this model brings prototype design and research center stage to encourage revision. In this model, stages of development overlap to illustrate that testing and refinement are an iterative progression. The instructional design process is “rapid” in that it allows for quick user feedback, while reducing the overall development time (Tripp and Bichelmeyer 1990). Research in rapid prototyping has led to the development of several adaptations (Batane 2010; Desrosier 2011; Dorsey et al. 1997; Piskurich 2011), including the more linear ID2 development model (Jones et al. 1992).
Fig. 2

The rapid prototyping ISD model (adapted from Tripp and Bichelmeyer 1990)

Although rapid prototyping is not a model frequently discussed in instructional design literature, it is uniquely suited for systems-based technical environments (Edmonds et al. 1994) as development on a project can begin before foundational research and design are complete. Furthermore, active participation from designers and students throughout the process is essential to shaping the final product. This ultimately increases the opportunities for responding to feedback, which in turn decreases the overall development time (Jones and Richey 2000). In such a fluid situation, designers using this model must stay focused on a central vision in order to maintain progress (Tripp and Bichelmeyer 1990).

The five stages of the rapid prototyping ISD model, assess needs and analyze content (AN & AC), set objectives (SO), construct prototype (CP), utilize prototype (UP), install and maintain system (IMS), were important to the design and development of TransGen Island. Over a two-year period, the many moving parts of this project aligned closely to the rapid prototyping model (Appendix 1). The next sections of this case study review each stage of the model in detail. We begin with the technical system for the simulation.

Install and maintain system

Although the technical infrastructure stage is presented at the end of the model, it must be front and center as an initial consideration when developing a technology-intensive project. What technologies are actually available can be a limiting factor when working in emerging technological fields. Technical specification, the capacity of available computers, and user experience have been frequent topics of discussion during the evolution of this project from the initial grant proposal planning up to the current day. Virtual world technologies are still in their developmental infancy with new innovations emerging daily, making the toolkit for this project a fairly dynamic one. Such dynamism both demands and allows for a great deal of creativity on the part of the developers and a model like the rapid prototyping model can be a true asset for maintaining control.

The project design originally included the use of Second Life as the platform for the simulations. In the fall of 2010, the team made the decision to move the simulations to an open-source virtual world. Second Life had removed the education discount, educators were choosing to maintain their own grids, and we experienced various technical limitations. A hosted instance of Open Simulator (OpenSim) met our needs for technical support while allowing us to maintain control of the server.

In spring 2012, infrastructure issues resurfaced during an apha-pilot within a university science course. The students in that course encountered immediate issues when several entered TransGen Island at the same time, therefore overloading the server. At this point, it was clear that we needed more control; we needed additional server capacity to accommodate a greater number of concurrent avatars, and we needed more specialized technical expertise. Some of our concerns could be addressed by hosting the software at our university, some could be handled by moving to the Cloud with a different server configuration. The immediate solution included increasing the RAM of the current hosted server to allow for as many as 30 simultaneous avatars, but a future plan will need to be developed to allow for significantly greater number of simultaneous users in several courses at one time.

Assess needs and analyze content

The team engaged in an extensive process of conceptualizing and revising our project, beginning with meetings between team members and science faculty individually and in small groups. We began our needs analysis by understanding individual science faculty backgrounds and their pedagogical needs. We were then able to synthesize our goals and brainstorm with groups of science faculty in working group sessions meant to crystallize the simulation design elements.

Initial meeting

Prior to submitting the grant proposal, we needed to come to an understanding of the major obstacles to creating a culture of philosophical reflection on ethics within the science curriculum. The obstacles included the ostensible value neutrality of science and the commonly assumed subjectivity or relativity of ethics. Although the meetings fostered interesting conversations, participating science faculty were generally not inclined to acknowledge subjectivity in science or the comparable objectivity of ethics. It appeared that the language and concerns of philosophers and ethicists were unrelated to the goals of scientists. The unsatisfying results of this initial conversation led to a rethinking of our strategy, including how we would interview faculty, what our workshops would entail, and what ethics pedagogy the simulation would employ. Our team decided that we needed to incorporate science faculty input into the project design from its outset, rather than merely trying to force a fit between our concepts of ethics and their concepts of science as the project advanced. We could work from those conversations toward our vision.


Sixteen faculty members in five different science fields were interviewed about their exposure to ethics training and their perceptions of key ethical issues in their disciplines (2010 Feb and Mar, Appendix 1). The interviews gathered information around two central issues: (1) the extent to which faculty had undergone and/or now require of their students ethics training, including elements of the responsible conduct of research (RCR); and (2) the perceptions faculty have of major ethical issues facing their particular fields, both currently and historically. The results were shared at the first Faculty 1 workshop. A large majority (13 of 16) of science faculty in the initial interviews confirmed having little or no formal training in research ethics during their education and only a few (4 of 16) currently offer more than small units on proper citations and data reporting procedures. Equally important, the faculty reported a wide variety of discipline-specific ethical issues, current and historic, which they wished they had time to treat in their curricula. The interviews generated solid usable content for the workshop discussions, and also generated enthusiasm amongst our Faculty 1 participants, boosting their investment in our project.


Two days of working group sessions were held at the end of each academic year, aimed at accomplishing two main tasks: (1) exposing faculty to information on growing trends in RCR/research misconduct and to ethical frameworks for discussing cases, and (2) drawing out attractive trans-disciplinary science content and relevant trans-disciplinary ethical decision making situations. Together creating the virtual reality environment that would be most appropriate and workable.

The results were exciting. While many good ideas emerged, we gravitated towards ideas that could use a single virtual environment regardless of discipline. The group decided to stress the importance of interpretation of research data, which speaks to concerns across all disciplines and to some of the gravest research misconduct. Students would be tasked to sample the environment and graph their results. The first ethical situation would arise when the students would be prompted by their virtual project director to sway their interpretation of the results for some reason of expediency. In light of Helton-Fauth’s (2003) findings about the need for ambiguity, making the request as reasonable as possible while leaving the door open to a moral judgment call was imperative for our goals. The focus of the player in our virtual world would be on completing assigned tasks and not on offering up a supposed “morally right” choice. Ethical reflection would be encouraged after the student had made his or her choice in an ethically ambiguous context.

One-on-one meetings

To transform conceptual ideas into the computer simulation, select faculty and student assistants met individually with the design team (2010 Oct, Appendix 1). This allowed for deeper exploration of the scientific and ethical issues as related to technical limitations. We eventually narrowed our options down to one discipline and one faculty member’s research expertise, looking for the most efficacious route to our first simulation.

Set objectives

The logical next step was to combine the information from the interviews, workshops, and meetings, with the best practices in science and ethics. This translated into “design choices” rather than formalized learning objectives. Although clearly-stated learning objectives are a favored method of instructional design, that approach limits creativity and changes in direction during the technical design process. The pedagogical goal was for the simulation to address the particular ethical issue (i.e. interpretation of research data) in an authentic context (i.e. under pressure from a superior). In addition, the experimental nature of the OpenSim software made some of the technical design decisions difficult. A declaration such as “students will collect and graph data in-world” can be problematic when technical limitations preclude certain outcomes and impact learning. For example, three of the design choices are listed below:

Design choice: delayed ethical evaluation

Student users are not prompted to consider stakeholder investment in the research outcome until after they have completed the simulation. This design choice was made to ensure that students gain the most insight from their experience. Students who are primed to consider ethical dimensions will often attempt to second-guess the instructor’s intentions rather than engage in authentic deliberation (Helton-Fauth et al. 2003). Our goal was to create a situation ambiguous enough that preconceived notions of “the right choice” could not interfere with students’ genuine thought processes. Delaying their reflections on and evaluations of the ethical dimensions of their choices by creating ambiguity in the assignment minimizes the possibility of students second-guessing their decisions or the intentions of the simulation. Since the virtual world in many ways protects students from the consequences of poor decisions, our team decided that our environment was an effective way to provide that opportunity to students.

Design choice: open ended

The decision students face includes elements of risk. For example, their superior suggests for them to adjust the data in a way beneficial to the project’s success, thus raising questions about the importance of the integrity of their data; maybe also about the moral integrity of their supervisor and the subordinate/supervisor relationship. As our science faculty (Faculty 1) taught us, science disciplines vary regarding the acceptance of the data points in question, and how much manipulation of data is consistent with accepted scientific practice. In order to see any problem with their supervisor’s suggestion that they massage the interpretation of their data, the student has to know enough about data collection and the standards of their discipline to be able to judge whether this is an appropriate request. After accomplishing their task and making their decision, students are posed a series of questions prompting them to reflect on their decision making process (e.g., how concerned would you be if these fish were to escape in the wild and mate with genetically pure salmon?). Later, in class or online, instructors can then take up a conversation about the students’ experiences and decisions. Instructors will then follow up on the simulation activity with a description of what the possible fallout of the various alternatives might be. As the project advances, various consequences of particular decision-paths might be built into the design, but presently we favor the flexibility and manageability of leaving the case open-ended. Students can come and go as they please once they have created their avatar giving them the choice to explore the simulation again. There are other areas of interest (e.g., a Hall of Fame for Science Ethics) in this open environment, and students can wander at will, allowing for individual follow-up and reflection beyond the assignment.

Design choice: Hall of Fame

The Hall of Fame is a portion of the TransGen Island activity that is not formally assigned but, like the gift shop in a museum, the flow of their activity will place them in its vicinity as their activity concludes. Curious students are free to browse through this space where we have accumulated stories of real researchers who have exhibited exemplary ethical decision making; scientists who have, despite the potential for career-ending actions in some cases, chosen the moral high ground. Too often our discussions of ethics in science emphasize examples of the most egregious misconduct by researchers. Providing examples of exemplary behavior in conjunction with the students’ authentic decision making experience encourages students to contrast their own experiences with the ones they are now reading about. This gives them further opportunity to reflect on the ramifications of the choices researchers must make. Such reflection plays a role in understanding the norms of one’s discipline and that, over time, has an impact on an individual’s sense of self (Hannah et al. 2011; Tice 1992).

Construct prototype (design)

In the first year (2010–2011), the development team was formed with three advanced undergraduate computer science majors. Mid-year, the students independently determined that they would rather have 2.5 computer science positions and .5 graphic design position. Their work resulted in over 2,000 h of programming time during the first academic year. This was increased to four full-time positions during the 2011–2012 academic year.

With a development team consisting mainly of undergraduate students, we paid careful attention to the development of technical skills and teamwork. The first 1–2 months of each academic year was spent learning the scripting language, visiting other virtual worlds, and creating fun objects (i.e. jetpack, avatar cannon, hole digger). This process was repeated at the start of each academic year to give new student assistants the same opportunity for growth. The team also worked on complex scripts included a wearable object to collect data, a receptionist who can give avatars access to restricted areas, and a quizzing tool. In addition, the students conducted a series of “show and tell” events demonstrating special projects to the faculty research team. These experiences provided students with responsibilities and roles similar to the 21st century workplace, an important aspect of the NSF proposal. That advanced computer science students were participating throughout the project meant that a secondary effect of this project would be that they too would have an opportunity to reflect on the importance of ethics in the sciences. While they were making both virtual jet packs and virtual posters with questions designed to promote ethical reflection in the student users, the process would ideally have the same effect on the programmers. Building the environment gives the programmers an experience consistent with the goals of the larger project.

The weekly team meetings of the student programmers and the supervising faculty member kept the students focused and engaged while providing opportunity for the supervisor to assist in problem-solving. Student programmers were able to divide labor according to interest, but they gave each other clear feedback when design contributions did not function “just right.” For example, the scripting involved in creating a virtual spreadsheet system for the simulation turned out to be physically do-able, but in reality too glitchy to be practical. Student assistants took turns tackling the scripting problems for over a year (May 2011 to Feb 2012, Appendix 1) before the decision was reached to remove that particular element. The outcome resulted in the need for students assigned the simulation to collect data within the virtual environment and then use a traditional spreadsheet software (e.g. Excel) to graph. So, another underlying effect of this project for the student programmers, while not formally assessed, was invaluable exposure to the importance of cooperative team efforts.

Utilize prototype (research)

The utilize prototype phase of the rapid prototyping ISD model is essential in determining the effectiveness of learning objectives and design choices, as well as providing feedback on implementation. Over a one-year period, the project team researched the implementation of the prototype with a focus on student and instructor experience as a result of the simulation. The research began with team members’ testing of the simulation, and expanded to include the participating faculty (Faculty 1 and 2), alpha testing within a small classroom, beta-testing in other universities, and finally follow-up interviews with instructors who beta-tested the simulation in their courses.


This study included 53 students from four different courses in Spring 2012 taught by three different instructors. Student participants attended three different universities, two in the USA and one in South Africa, and they were enrolled in a range of courses (science, engineering, and computer science), degree programs (undergraduate and graduate), and instructional delivery methods (face-to-face, blended, and online).

Data collection

With the module built, student data collection occurred from February to May 2012. After students completed the assignment, we examined survey data, student reports, and location data. The 16-question survey included six questions on the individual’s experience in the virtual environment and asked the students to self-report on whether they acquired knowledge in science and awareness of ethics. In addition, other questions provided feedback to enhance the remaining revisions in the design process. Data analysis included descriptive statistics and Pearson correlation coefficient of survey data and review of student reports for emerging themes in science content knowledge and ethical understanding.

We also conducted one-half to one-hour long qualitative interviews with the three instructors who participated in the beta-testing from May 2012 to July 2012. Interviews were conducted and recorded by two team members via Skype. The interview protocol included questions on academic background and training, current experiences teaching ethics in their courses, and implementation of the SciEthics Interactive simulation. Participants’ responses were coded into emergent categories (e.g. “student technical problems”; “time spent preparing for activity”), and analyzed thematically.

Student results and discussion

Students had a positive response to the simulation, indicating changes in science knowledge and ethical understanding. Students reported an increased awareness of ethical issues, recognized the ambiguous nature of the data, and concluded that their findings were important to society (Table 1). The majority of students reported increased learning of the science of genetic modification (Fig. 3) and their responses showed a significant, moderate correlation (.453, p < .05) to the learning of ethical issues surrounding genetic modification (Q6, Table 1). Students who reported learning more about the science issues were also more likely to report learning ethical issues surrounding GMO’s. Survey question 2 and 3 also showed a significant, moderate correlation (.505, p < .05), indicating students believed the knowledge gained was valuable to society. Overall, the students spent an average of 3.7 h exploring the simulation. The majority of students completed the in-world portion of the simulation in less than 5 h.
Table 1

Student survey results on science knowledge and ethical understanding

Question no.

Question text





The results of the data were ambiguous





The data in this report is scientifically valuable





The knowledge generated by your report is important to society





I learned something scientifically valuable about genetic modification





This exercise has been worthwhile from an educational standpoint





This exercise raised my awareness of ethical issues around this research




N = 53
Fig. 3

I learned something scientifically valuable about genetic modification. 1 = nothing, 2 = a little, 3 = some, 4 = a lot

The analysis of the written reports confirmed that students did recognize the ethical issues surrounding the genetic modification of salmon. Students reported an average of three (m 2.77, SD: 2.052) ethical issues with a range from zero to eight. The ethical issues of environment (67.9 %), salmon health (56.6 %), and human health (54.7 %) were mentioned in the majority of student reports. Although the instructions for the report did not ask students to reflect on research ethics, 15.1 % of the reports mentioned an inadequate amount of data and 17.0 % of the reports mentioned bias from the supervisor (Table 2). Illustrating the ambiguity of the simulation, students’ final recommendations for human consumption of the TransGen salmon were split between no recommendation, no approval, further study, and approval (Fig. 4).

The student reports clearly indicated a significant reflection on the issues we were trying to raise, and they took the activity seriously. Amongst these initial participants, we see the following themes emerging from the research reports: (1) the students absorbed a large portion of the facts they were exposed to and (2) they acknowledged that the ethics of the activity mattered. This is an important first step in understanding the impact virtual worlds could have in the academic setting.
Table 2

Student report excerpts on ethical issues at TransGen Island


TransGen salmon will be raised in fish farms that, even when raising native breeds, cause significant environmental damage. They essentially operate as underwater factory farms, with thousands of fish swimming in their own waste, which in turn requires high doses of antibiotics. The fish are prone to escape, and when they do, they compete with wild fish for food, and they breed with those fish, slowly causing wild fish populations to become extinct

Salmon health

Whether the local media finds out the issues that are happening at TransGen or not, we need to voice our opinion to the authorities regarding the environment that those animals are living in, and the consequences of GMO foods

Human health

The biggest issue I have is that while this operation has been running for some time now, and TransGen claims to have complied with all FDA regulations thus far, it’s still too early to be able to definitively claim full understanding of all health risks. Often the FDA is under enormous pressure from industry and therefore honest mistakes may take place


Data available is still relatively small, and may be skewed, therefore I would continue to approve this GMO product for customer consumption, but I would do some in limited supply and continue to monitor the species and associated data

Supervisor bias

Regardless of the pressure put on me by my supervisor, I definitely believe that TransGen does not pass evaluation as a food product…Ultimately, my sworn ethical integrity is what has driven me to uncover the reality of this situation. My ethical concern is backed by the proven evidence given which directly proves that GMO salmon is more detrimental than advantageous
Fig. 4

Student recommendations for human consumption of salmon produced at TransGen Island. 0 = no recommendation, 1 = no approval, 3 = further testing needed, 4 = approval

In terms of the practicality and ease of use of the simulation, we learned a great deal from the student beta-testers. Moving and learning in a virtual environment was new to most students and frustrating to some. One student reported that “I couldn’t get the avatar to work correctly and the experience was not a good one. I would have much preferred doing a report had I had that option up front and with time to work on it.” Another student worked through some difficulties, but found value in this new experience:

I believe the SciEthics definitely pushed me outside of my comfort zone, which is something that I should expect when I am involved in higher education. I now realize the importance of taking on challenges that do not necessarily have importance “in the moment” but are intended to open me up to new experiences. These experiences add to my collective experience in spite of the discomfort that was created in process.

Within the structured activity, students reported varying degrees of comfort with the technology and time spent in the virtual world, indications of the personal factors that impact learning in virtual environments.

Instructor results and discussion

The follow-up instructor interviews were designed to examine the process surrounding implementation of the simulation as well as gather information about the perceived value of the activity for student learning. All three instructors entered the virtual world with their students, but to varying degrees; one (instructing one of the online graduate courses) made herself available in-world for several hours a day during the period students were assigned to complete the simulation, while the others explored the island and made themselves available periodically to answer questions about it. Moreover, the instructors had access to all the students’ work on the assignments and the self-reporting questionnaires students completed. As a result, they were able to comment on students’ perceptions. In broad terms, they affirmed much of what students reported: the simulation was worthwhile and informative, but some students found it a little cumbersome and/or unnecessarily game-like. However, all three instructors contended that nearly all the students got more from the experience than the students themselves self-reported, a claim that is difficult to assess.

The interviews show areas of substantial agreement among the three instructors in terms of both positive and constructive feedback. They report general satisfaction with the simulation and their experiences using it. Students showed clear evidence of having learned the concepts intended in the simulation. Two instructors were careful to point out that they had already determined that to be the case before they looked at the students’ self-reports. One instructor, who had used a different virtual world activity in a previous course and described it as “stupid” and “without purpose”, was skeptical about the value of virtual worlds. Her view changed after implementing the SciEthics Interactive simulation, emphasizing that “with this one, pedagogically, it was beautiful.”

Further, all three instructors, when asked about possible improvements in either the simulations or the supporting materials we provided, reported that they would be ready and willing to use the simulations again (regardless of future improvements). Two respondents actually used the phrase “I will” in response to a question that asked conditionally “What would you do differently?” The third respondent, in response to another question, indicated that he already plans to implement the simulation elsewhere in his program’s curriculum, where he thinks the self-motivated nature of the experience and differing levels of students’ identification with their avatars can better serve their needs than by building it into a specific course.

We received constructive feedback from the instructors about how the activity is framed in the supporting materials for both instructors and students. All three instructors recommended more information for the students about the mechanics of developing and navigating their avatars in-world. This feedback is consistent with the student self-report data provided above. Because students spent longer than expected simply learning how to move, grab items, find objects and so on, the instructors also recommended indicating, in the instructor manual, the necessity of providing enough time at the beginning of the assignment to allow for that learning curve.

Similarly, the instructors agreed that they would like more recommendations for themselves about how and when to use the simulations in their courses. Because they were beta-testing, we had instructed them to use the simulations however and whenever they thought best, and hence they reported that we should make clearer and stronger suggestions about how new users should situate the simulation within courses. One respondent suggested adding list of topics for class discussion before the activity; another suggested advising instructors that they encourage students to work in groups while in-world.


The design, development, and data collection process for TransGen Island offers several important lessons and recommendations for anyone embarking on projects using 3D virtual software. As we hope that our work will inspire others to take on similar projects, we have accumulated a list of “lessons learned” when creating virtual world simulations.

Lesson 1 Formally checking in with instructors who assigned the simulation and getting feedback from all users is an integral part of refining a finished product. In our case, as noted in the interviews with instructors who beta-tested the simulation, more instructional support would be welcome. We knew from the beginning that asking faculty to add more content into their already busy courses would draw some resistance. We recognize from the instructors’ interviews that they would have liked more help making pedagogical and curricular decisions. Put more broadly, the multitude of ways in which this project relied on the participation of others was key to its success. Maintaining collegial connections with diverse corners of the home university and other participating institutions was essential.

Lesson 2 Identifying ways in which students acquire their “in-world” skills would have assisted us to identify more helpful suggestions for the manuals and changed aspects of the world as it was being built. While we made it a priority to ensure faculty buy-in, and thus focused on acquiring feedback and then commitments from them, we can imagine a student-oriented working group could also have been a valuable and productive route for gathering initial feedback. Our experience was, however, a fairly smooth transition from faculty input, to creation of the simulation, and then finally student feedback at the beta-test phase.

Lesson 3 Getting science faculty to commit to an ethics project is an absolute necessity for a project as this. That in itself accomplishes an explicit mission of the NSF’s EESE grant monies. Faculty who appreciate the need for ethics education will get ethics communicated to their students, thus propagating the value of ethics training into the next generation of researchers. That said, not all science faculty are going to be committed to those kinds of goals. As discussed earlier, the first level of generating science faculty interest was to set up a series of interviews with science faculty asking them about their professional development and the level of exposure they had to the language of ethics and to ethics education. During that process, enthusiastic faculty could be immediately identified. Targeting that group for participation in the working groups successfully filled our workshops. We also offered a monetary incentive and assured those still in the tenure-track pipeline that letters of support for their tenure applications would be gladly forthcoming (all those faculty in that pipeline subsequently did request letters).


The results of this research project raise three general themes regarding the use of virtual environments for student learning: (1) the importance of interdisciplinary collaboration, (2) the value of teaching ethics in virtual environments, and (3) the use of rapid prototyping as an instructional design model. First, this project illuminated the importance of connecting with content-area faculty when designing educational simulations. The interviews and workshops helped the team clarify both the need for particular forms of ethics education and science faculty’s desire for an effective way to teach them. The issues our simulations raise emerged directly from their responses. Faculty members’ interview responses indicated a clear gap between the ethical issues arising in science fields and ethics pedagogy for science students. The existence of the NSF Ethics Education in Science and Engineering program highlights this gap and seeks to respond to this need on a large scale (Hollander 2009). Our research confirms that this is an important issue for faculty, and our project offers a useful and efficient contribution for filling this gap.

Second, the students who engaged in this virtual world experience were exposed to ethical issues they had not explored before. They were given a chance in a freestyle environment to raise questions and ponder issues textbook learning can rarely give. Students reported a positive experience with the simulation and an increase in ethical understanding. For instructors who are not well-versed in the language of moral philosophy, TransGen Island is a valuable alternative to bypassing or inadequately treating questions of ethics that arise in the process of practicing science. Our project offers a concise ethics primer to help prepare instructors, and more importantly provides an open-ended experiential activity for students to explore the ethics of research in community. Instructors do not need to supply the “right answer” so much as simply be willing to facilitate conversations about the questions raised in the virtual activity. Since our virtual-world simulation presents ethical-decision making in the conduct of research as it would occur in practice, it invites instructors and practitioners to draw on their own experience in their professional life for comparison and exploration.

Third, the rapid prototyping instructional design model served as a strong foundation to the development process of this virtual simulation. Each stage in the 1990 model figured in the creation of the final working product with a slight modification for the initial consideration of software and hardware. As an overall instructional design method for virtual environments, rapid prototyping deems appropriate. Additional research is needed to determine the extent to which the rapid prototyping shortens overall development time in virtual worlds. Even with these challenges, the rapid prototyping model will be a useful planning tool for any academic instructor planning their own 3D virtual environment projects.

The SciEthics Interactive project had a positive effect on the learning process for students, in large part, because of the exposure to ethical reflection from which many individuals might benefit. Future research is needed to examine the unique characteristics, attitudes, and behaviors of individual students using virtual environments for learning activities. Further research is also needed on the virtual world’s role in ethical development over time. As indicated within this study, the student experience can vary greatly. Some students embrace gaming technologies, while others may resist learning in immersive environments. To broaden the reach of this project to a global audience, the island files and documents are publicly accessible on the internet at


This research was supported by the National Science Foundation EESE Program Award No. 0932712.

Copyright information

© Association for Educational Communications and Technology 2013