Conducting research in virtual worlds

Texto da comunicación

Introduction

The virtual worlds, which have been democratized over the past decade, nowadays represent high economic and social stakes. This paper presents methodological and ethical elements that researchers may take into account if they want to conduct a project in cyberspace. The emergences of virtual worlds on the Internet and their quick acceptance by society have led to the rapid development of technologies related to virtual reality (Castronova, 2005). Consumers (El Kamel, 2009) and companies have not hesitated to visit these new online spaces where everything seems possible. By the end of 2011, 80% of the Internet users will have a second life online (Gartner Group, 2007). Among the virtual worlds, Second Life is one of the most popular with several million accounts created since it was opened. Users can find at anytime between 35,000 and 82,000 other people connected all around the world. According to the Official Linden Blog of Second Life (Second Life Official Blog, 2009), users spent in 2008 nearly 400 million hours of attendance in Second Life (61% more than in 2007) and transactions between them reached a value of $ 365 million US (33% more than in 2007) Because of the growing importance of these virtual worlds, economical, societal and legal questions are appearing. As a consequence the Organization for Economic Cooperation and Development (OECD, 2008) has started to raise issues such as the rights of intellectual property, privacy or identity and more generally topics about the regulatory problems arising from the increasing prevalence of virtual worlds. Like any new environment, Second Life and other virtual worlds provokes fears and doubts, but the increasing involvement of Internet users (USC 2008) and the arrival of a "cybergeneration" (2002 -2025) are factors that suggest that this phenomenon will continue to grow quickly over the next decades.

Many companies see marketing opportunities in the virtual universes (Manu, 2007) and both departments of research and development and marketing have entered or are integrating them. Hendaoui, Limayen and Thompson (2008) speak about the emergence of a new paradigm: the virtual marketing (v-marketing). The use of virtual worlds provides new questions for companies and researchers:

• What are the new opportunities offered by these metaverses?

• What types of research methods can be implemented?

• What are the constraints that can influence researchers who wish to work in this new type of environment?

• How to adapt ethical standards in virtual worlds?

• etc.

Presentation of the Second Life metaverse

Second Life differs from traditional video games. More than a game it is a virtual environment, a parallel world (Second Life, The Official Guide, 2007). Users (usually called residents) have no predefined goal or quest; they can spend time freely in this virtual environment and live, virtually a second life, at their discretion. For this main reason we talk about Second Life as a metaverse (Stephenson, 1992). Residents can try to represent themselves as they are in the physical world or create an entirely different character. For example, they can take the appearance of an object (like a robot) or of a fictional animal (as a furry). The "land" where the avatars can freely live consists of three major continents and a multitude of islands in the intercontinental oceans (30,000 regions in August 2009) originally designed by Linden Labs. Residents can buy or rent islands or plots which are measured in square meters and associated with prims, the standard units in Second life. The monthly costs (called tiers) to maintain the properties or leased land on the servers of Linden Labs are determined by the number of prims that a land can contain. Typically a plot of 512 square meters contains 117 prims.

Second Life reproduces several aspects of real life. So, it is characterized by the presence of an economy based on virtual currency, the Linden dollar (L$). These dollars can be bought and sold on the Linden website like a real currency and the avatars can spend their money in several products and services such as clothing, accessories, sporting goods, real estate or recreation. Second Life is a consumer society in its own right even if laws from the physical world can interfere with the development of the virtual community. For example, European users have now to pay VAT on their transactions with Linden. Objects, commodities or services, everything is virtual, implying new ways of consumption, hence the interest for many companies and researchers in marketing. Among residents’ activities we find also creation, selling and trading of virtual objects on which they can keep their property rights. The social aspect is very developed in Second Life (Ludlow and Wallace, 2007) actually avatars can interact freely with each other, typing chats and instant messaging (IM) or speaking. They can also join groups or communities pursuing common goals or interests. Whatever the motivations (to meet people, share a passion, live out their fantasies, etc.), Second Life is a real second life for many residents. This aspect of this virtual environment is admirably described by Lucie Dos Santos (2008).

Many companies have begun to integrate Second Life during the past few years. Investments have already exceeded millions of dollars. For example, IBM plans to invest ten million of dollars in the virtual worlds in the coming years. Marketing has also its place in this virtual environment. Brands like American Apparel, Lacoste, Reebok and Toyota offer and /or test their products in different stores for communication or sales purposes. The enthusiasm of companies is not limited to advertising. Human resources also find their place in Second Life. For instance, companies such as Accenture, Alstom, Areva, Capgemini, L'Oréal and Unilog use virtual worlds for recruiting their employees.

Characteristics of virtual reality: research perspective

Virtual reality refers to technologies that provide a representation of interactions and a multi-sensory presence by combining real-time simulation techniques and managing the virtual environment (Amditis, Karaseitanidis and Mantzouranis, 2008). Virtual reality, as Second Life, is now accessible to the general public but we must point out that applications of these technologies have been used in the product creation and development for several decades. One of the most significant partnerships was set up in 2004 between the French company Dassault Systems and the U.S. firm Boeing, which created the airplane Dreamliner with a platform for virtual product development. The democratization of technology using virtual reality is in large part linked to the proliferation of online games such as the massive multiplayer online role-playing games (MMORPG) World of Warcraft (WoW) or the Sims. The term virtual reality is ambiguous because it refers to various environments. Nevertheless, their common features are the free evolution of avatars and the freedom given to them through the interaction human-machine (residents in Second Life). In a MMORPG such as WoW, a player participates in a virtual world but he must accept the features authorized by the designers of the game. On the social network Cyworld, which has more than 60 million members, they do not evolve in a virtual world itself but they have the opportunity to create a virtual avatar. Amditis, Karaseitanidis and Mantzouranis (2008) distinguish two categories of variables to consider for conducting research using virtual reality. These categories refer to aspects related to the content available in the virtual world (content aspects) and the tools available to interact in this world (tool aspects) Each of these two categories consists in a set of six variables (Table 1) which may have potential opportunities but could also be barriers to the achievement of scientific research.

Table 1: Variables directly related to virtual reality and to consider in a research perspective.

Variables directly related to virtual reality and to consider in a research perspective

1. Content available in the Virtual World

1.1. Presence

Among the different variables related to the content offered in the virtual world (content aspects), one of the most discussed refers to the presence that users may experience in this type of environment. The concept of presence focuses on two distinct aspects in the literature. The first aspect concerns the immersion or absorption that the virtual world allows the user to feel. Indeed, online experiences like experiences in real life, can be immersive with the help of technologies (Mathwick and Rigdon, 2004) and users can even experience a state of flow (Csikszentmihalyi, 2002). Virtual reality facilitates this immersion and, in Second Life, this immersion is more effective than in a traditional online environment. Second Life can be regarded as a metaverse (Stephenson 1992) because the presence of individuals is represented by avatars. Participants can feel someone there both "physically and literally ..." (Stewart and Williams, 2005). Although Second Life is in many respects an environment that reproduces the reality that one can know in the world "physical", other environments less "realistic" (The Sims) or even fantastic as World of Warcraft allow to users to experiment immersion in a virtual world. We must take into account that immersion is promoted by features of the virtual environment and their use will depend largely on the involvement of the user into the virtual world. Because there is not a clear paradigm for the concept of immersion in the virtual universe, the concept of presence is still in question by its detractors.

The second aspect is the ability of avatars to regroup in the virtual worlds which we refer as a social presence. Thus, a benefit for the researcher (the most commonly cited about conducting a research in such environments or with methods of online research) is the possibility of having the world as a big field of research (Witmer et al 1999). In the case of Second Life and other virtual worlds this advantage has a second dimension that consists on the opportunity to study all metaverses for themselves. It means that the purpose of the research can be focused on Second Life not only as a way for researching but as an end in itself. The group formation is facilitated in virtual environments and today we find many ways to regroup avatars and to assemble the sample for the research. This can be done thanks to the fact that there are many interesting groups in Second Life where a researcher can be in contact with their members. The contacts are also facilitated by technology: a researcher can communicate with an entire group by posting a message on a forum dedicated to that group or by contacting an avatar using private instant messages. Second Life has both the metaverse and the social presence properties which facilitate the users’ immersion and their feelings of presence.

1.2. Shared spatial experience

The second variable related to the content offered in the virtual world is the spatial experience shared by all actors. In the case of Second Life that offers almost unlimited flexibility to users, the spatial experience shared by all players is represented by the units of measurement which are provided. The prim is the base unit to design an object and represents the smallest common denominator for all avatars. The terrain is another unit of measurement for builders and the Linden is the official currency. Towards these elements, the designers of Second Life provide a language and a common area where all players can evolve. Jarvis (2009) stated that “enabling customers to collaborate with you [a firm] - in creating, distributing, marketing and supporting products- is what creates a premium in today’s market”. As a standard unit, primitives enable residents to create whatever they want. Thus, Second Life offers an undeniable advantage compared to video games: openness.

1.3. Experience, 1.4. Knowledge, 1.5. Understanding

These variables are closely linked together. They refer to the learning, familiarisation and control of the virtual environment by the user. Whatever the virtual world where a researcher wishes to carry out a project, these variables must be considered carefully because they may have an impact on the conduct of research. For instance, we must consider if the research method selected requires sustained interactions between the various parties involved. In a virtual world such as Second Life, learning and familiarisation by the user are easily accessible as evidenced by the large number of users. Tools available for users in order to communicate or travel in the virtual environment have a crucial importance to facilitate interaction. In Second Life, there are different modes of communication available: private messages, forums or oral communication and modes of travel (walking, flying and teleportation) but also tracking the position system. One potential inconvenient for researcher is the necessity to have a learning phase so as every participant achieves a minimum level of environment control. For this reason, recruiting participants outside Second Life to do a study in the metaverse implies an additional difficulty for the researcher.

1.6. Creation of a Story

The final variable related to the content offered in the Virtual World is the creation of a story. Generally, the creation of a story is defined by the designers of the virtual world: it provides user guidelines which have to be followed to evolve within the virtual world, like the "missions" that players must achieve in World of Warcraft. In the case of Second Life there is no a predefined story provided by the creators of the metaverse. This allows residents to create their own history. This freedom is one of the reasons for the success of this metaverse but, from a research perspective, this freedom has a dual effect: first, positive because the researcher will have a great diversity in of avatar lives, which can be a source of information richness, but also negative because the criteria for obtaining a homogeneous sample will be difficult to establish.

2. Tools to interact in the Virtual World.

2.1. Interaction

In the second category of the variables proposed by Amditis, Karaseitanidis and Mantzouranis (2008), the interaction between the user and the virtual world is a key variable. Harvesting information via qualitative methods such as individual or collective interviews in virtual worlds would offer opportunities for richer interpretation of the data (especially in comparison with the interviews conducted online). Several reasons have been advanced to justify the superiority of virtual worlds in relation to online and offline methods. The first one is based on the fact that the use of avatars could facilitate exchanges and discussions between the participants but also with the researcher. Although several studies (Colburn, Cohen and Drucker, 2000; Freier, 2007) have shown that there are still limits concerning communication with avatars, Stewart and Williams (2005) explain that the closeness between the avatars and the movement of the body can facilitate the interpretations of the researcher. Indeed, Goffman (1959) emphasizes that in social situations, individuals offer a partial and controlled attitude so this may have a negative impact when the researcher wants to study a phenomenon in detail. Thus further tests such as body language can help the researcher in its interpretation. Mc Arthur (2008) states that such a phenomenon is possible in a metaverse like Second Life. Nevertheless, in Second Life it could be difficult to interpret (especially the body language of avatars) because of a delay between the gesture request by the user and the avatar movement in the virtual world can lead to a false interpretation.

Another argument valid not only for online research but also for research in virtual worlds, is the ability that the researchers have to broaden the interpretation they do with the numerous abbreviations used by the participants such as lol (Laughing Out Loud) or hmm (indicating a reflection) that are "mood" indicators helping for the interpretation (Sternberg, 1998). One double edge situation in a metaverse concerns perceived anonymity by the participants (Joinson, 1998, Stewart and Williams, 2005). Due to the perceived confidentiality of data, some participants could reveal more about themselves than they really want, which raises ethical issues for the researcher. However it gives extensive data which could help the researcher to explore in details a phenomenon. This last point highlights the potential of virtual worlds where communication being with the avatar can be viewed as an intermediate solution between the strong perceived anonymity in online research and the information retention presented by Goffman (1959).

2.2. Spatial interface

The second variable related to the tools used to interact with the virtual world space is the spatial interface. This one is particularly important for researchers because it defines the range of techniques that they can use to achieve their work. Virtual environments have at least the same features than online environments, allowing oral and written communications, but also allow exchanging documents such as photographs that can serve as cornerstones during interviews (Loeffler, 2004). Another opportunity provided by Second Life is the fact that an object can be easily modified by an avatar. This will allow a company to obtain, at a low cost, the point of view of the consumer about the product design, for example. Similarly, Second Life gives architects the opportunity to create models of their projects where participants can walk and share comments with their creators. In this way, the interface space in virtual worlds, especially in Second Life, appears to be a field of research more flexible than both online and physical spaces. Therefore it offers many opportunities for designing and testing new products, even for developing a brand image by using virtual marketing (Hendaoui, Limayen and Thompson, 2008).

During the interaction between the researcher and participants, respect is a key issue. As a consequence, in this kind of environment the researcher should not impose a particular language: he/she has to make the effort to adapt to the context. He/she must not require a different mode of interaction, especially if the person prefers to write rather than to interact verbally. The researcher must also manage the time for discussion by giving the avatar the time required for the interview or for participating in the discussion. But all the respondents have not the same rhythm when typewriting a chat or an IM. It is obvious that the faster ones will tend to monopolize the session in the case of a focus group. For that reason, the researcher should inform the interviewees of the length of discussion and should also provide enough time to record the responses of respondents and give them the opportunity to interact. The typewriting pace should not be either frustrating or a brake for the avatar.

2.3. Assessment, 2.4. Planning, 2.5. Development, 2.6. Creating a situation.

Assessment, planning and development variables refer to what must be put in place to pursue research in a virtual world. These elements depend primarily on the subject of research but it is important to remember that in Second Life we have a high range of possibilities. In this way, we can reproduce many real-life situations such as shopping or meeting in places to conduct group interviews. Similarly, the last variable, creation of situations, is largely related to the previous ones and gives flexibility to the researchers giving them almost unlimited possibilities.

Variables indirectly related to Virtual Reality and to consider in a research perspective

Amditis, Karaseitanidis and Mantzouranis (2008) propose an analysis grid based on the technological properties of a virtual world. However, it is possible to identify other factors that may inhibit or encourage a researcher to use a virtual world like Second Life as a research field. One advantage of these environments is that they can benefit from a novelty effect on which a researcher can capitalize. Consumers are constantly solicited to respond to surveys and exposed incessantly to advertisements. Virtual worlds are still moderately affected by these phenomena and the researcher can find individuals deeply involved in these environments. Nevertheless, there are still several barriers to the use of virtual environments:

• Financial: because the efficient use of a virtual world is still expensive (users need a powerful hardware and a fast Internet connection)

• Technology software and hardware are not always easy to use.

These barriers are real but the democratization of technologies using virtual reality is reducing these obstacles.

Ethics in virtual environments: what reality?

Ethics in marketing research

While the importance of ethics is now widely recognized by the scientific community, the concept and meaning are rarely defined with precision. In Canada, one of the documents for the academic community is one proposed by the Social Sciences and Humanities Research (SSHRC, 2005) entitled "Ethical Conduct for Research Involving Humans" with eight guiding principles: respect for human dignity, free and informed consent, protect the vulnerable people, privacy and personal information, justice and integration, the balance of advantages and disadvantages, the reduction of inconveniences, maximizing the benefits. If these principles can guide the researcher in his quest, the meaning and relative importance of each of them have different interpretations. In marketing, the issues related to ethics have been discussed several times and some principles have emerged over the typologies proposed by the authors. Tybout and Zaltman (1974) have highlighted three basic consumer rights: the right to choose, the right to security and the right to information.

To a large extent, this first typology is grounded on the belief that the marketing image for a non-specialist audience is often negative. Helfer (2003) presents several examples that can explain why marketing has a negative image. According to the author, the development of a marketing database and the likelihood of being perceived as Big Brother are realities which imply many potential abuses. More generally, all components of marketing (advertising, product policy, quality etc. ...) are subject to trade practices to the limit of what could be considered ethical. However Helfer (2003) emphasizes that the right to choose is increasingly met with a consumer increasingly free with his/her choices (because more information is available). Also, marketing practices are more and more regulated. Holbrook (1995) proposes another typology consisting on four distinct categories of ethics in marketing research. This typology distinguishes two dimensions:

1. The first addresses the role of one who has to adopt an ethical approach (the researcher or the consumer)

2. The second focuses on the approach (reactive or proactive) in relation to the phenomenon studied.

According to Holbrook, a code of ethics in marketing must be interpreted depending of the four following situations: (1) marketing as a practice, (2) marketing as a relation developed with consumers, (3) marketing during the review process of academic journals and (4) marketing as a research field. We focus on this last situation for the following analysis.

Problems related to ethics in marketing research usually appear when the personal interests of the researcher are involved either due to the study topic or because of expectations in terms of results. It is possible to distinguish at least two major types of population that develop marketing research:

1. The professors

2. The consultants.

The results of research by these populations are usually addressed to different audiences, but in both cases there is an obligation of performance towards their employers. As Gendron (2008) highlights, the statement "publish or perish" is becoming even more important in the academic world. For the consultants, their mode of remuneration (hourly billing) and the obligations of profitability are all pressures that may lead researchers to give less importance to ethical issues in their research. Compliance with ethical principles and obligations related to these issues differ for these two groups of researchers. While professionals are generally limited to the principles of government legislation, in Canadian universities professors will have their research project scrutinized by the Ethics Committee of their institution. If this review is maybe necessary to avoid abuses, its represents an additional pressure for researchers, in particular because the review process is time costly. It is therefore justified to wonder about the ethical issues that a researcher can encounter if he chooses a metaverse as a field of study.

Major ethical issues specific to virtual environments

From an ethical point of view, research conducted in a virtual environment has some advantages when compared to conventional research because of the way in which participants meet the researcher. Due to the requirements of research, experimental procedures can lead the participants to want to withdraw for emotional or physical reasons (for instance, when the subject of the study can be classified as sensitive). In virtual environments, the participant is not physically facing the experimenter and, in many cases, anonymity is total. Thus, to exit from the research, the participant has to simply disconnect from the virtual environment, what can be achieved with a few clicks. However, in a conventional research, the participant must get up and leave the room under the gaze of the researcher and of the other participants. In this situation, the social pressure that weighs on the participants is much higher: it can impede the right of withdrawal. Studies on obedience to authority (Milgram, 1963) and conformity (eg Asch, 1951, 1956) suggest that the right of withdrawal could be "violated" if the research protocol is not adapted: participants undergoing too much social pressure would not withdraw from the study even if they wish to. From an ethical perspective, the use of virtual environments is suitable for researches on sensitive issues that allow participants to withdraw easily.

A second advantage of virtual environments in comparison with conventional research is the ability of these environments to reduce barriers for vulnerable populations, especially when that vulnerability is related to a physical disability. Indeed, in accessing a metaverse as Second Life, users are less vulnerable to many forms of discrimination (racial for example). In addition, young people, who are a vulnerable population with a strong presence in these virtual worlds, must receive a special attention in the sense that an avatar whose resident is under 18 years old (in real life) should generally be excluded from the search. To address this problem, designers of virtual worlds have implemented procedures to ascertain the actual age of the participants, as for instance, as entering a passport or an identity card number which can be easily checked on internet. Virtual worlds, in addition to provide researchers a worldwide study field, is also more accessible and, as a consequence, more fair and inclusive.

A third advantage of virtual environments in comparison with the ethical principles enunciated by the SSHRC, is the reduction of disadvantages. Conducting a research in a virtual world is often less demanding for participants who can participate in the experiment when they have access to an Internet connection and a computer. Similarly, opportunities to optimize the reward for participants are higher because researchers can offer compensation to them which will be directly sent to their avatar as Linden dollars in Second Life or can be sent to the user of the avatar (for instance, in real currency). If, in real life, the researcher can provide incentives to encourage the recruitment of respondents, one should remember that Second Life is a rich metaverse where "objects" that people would like to receive as a gift in exchange for their participation are numerous. The value of these objects (in Linden dollars) but also their nature should be specified by the ethical guidelines. Non consideration of the awards in the ethical guidelines for online research could be considered by the avatar as a disadvantage compared to real life or as diminishing the value of the identity of the avatar relative to his/her true identity. Moreover, this freedom of choice to individuals may be an additional motivation to be involved in the research.

Several ethical issues can be identified in terms of research within a virtual world. Clark (2004) identifies three principal issues:

1. the manifestation of the researcher presence,

2. obtaining the participant consent,

3. confidentiality and anonymity.

The manifestation of the researcher’s presence

The manifestation of the researcher’s presence is likely to raise problems of an ethical point of view, especially when observation is the method used because he /she has the ability to turn into a real Big Brother. In virtual worlds, it is possible to record with video capture programs all the facts and actions of individuals and this type of approach may, in research, become extremely intrusive. Also, it is necessary to put additional safeguards which do not rely only on the researcher goodwill.

Obtaining the participants consent

Obtaining consent can be more difficult to achieve in virtual worlds than in traditional research, but this difficulty is primarily technical. Indeed the main issue is to be able to recover and obtain the signature of the participant. Also, the researcher should make sure to implement a technical solution for the consent of the individual. Among possible solutions, the researcher may use a system similar to the one that companies use for selling software. They use a form on the Internet where the participant declares to accept the rules by ticking a box which will send a digital signed consent to the researcher.

Confidentiality and anonymity

This last issue has already been discussed when methods for online research were developed. The way research is conducted online and in a virtual world could lead participants to have the impression that their anonymity is protected because they do not appear physically to the researcher (Stewart and Williams, 2005). This perception may be wrong, and it is important that the researcher pays close attention to this and preserve the confidentiality of the data set.

This description of the major ethical issues, leads us to ask questions about the ethical rules governing research in metaverses, especially in Second Life. The avatar participating in a research must be protected by the ethical rules related to an environment in which things can operate sometimes completely differently from the real world.

Virtual environments and ethics: what researchers have to take into account

Internet today offers researchers a platform for expression of different forms of social life. The interest of these forms lies in their diversity, accessibility and persistence (Berry, 2004). Does research in such different worlds require an adaptation of existing codes of ethics with the aim of respecting the dignity of the participants? Some authors refer to the following as "golden rules":

1. Never mislead participants.

2. Never consciously put participants at risk.

3. Maximize privacy while minimizing potential harm (Thomas, 1996).

Reading these three rules, it seems clear that they are too general to be applied stricto sensu for research in virtual worlds. We must point out that online research, more particularly in virtual worlds, is in permanent change. Because of the changing nature of these environments, flexibility is also required as a remedy against the uncertainty (Sollie, 2007). Technology is the tool which gives us access to these virtual worlds. Uncertainty being inherent to a machine, it should be reduced by a greater flexibility given to the researcher. Therefore, it is quite natural that ethics follow these changes and adapt to them. Bruckman (2001) speaks in this context of “flexible ethics”. Each research is by nature different but in a virtual world, its uniqueness will also depend on the use of the virtual environment in the research process. We distinguish two major uses: (1) the research object deals with the virtual world or one of its components for itself, (2) the study is achieved in the virtual world which is only a new mean of action for the researcher.

For instance, the research object could be to:

• Investigate aspects of the avatars’ social life and their evolution in the virtual world.

• Study the social behaviour of the person behind the avatar.

• Investigate aspects of group memberships of the avatar…

When a researcher uses a virtual world as a new field of investigation, he/she could benefit from the technological advantages cited earlier. However, on an ethical perspective, another variable must be taken in consideration: the research scope. Depending on the study goal, the researcher would interrogate only the avatar, only the person behind his/her computer, or both of them. If one tries to imagine all the combinations that may arise between these two typologies (use of the virtual world and research goal), we realize that ethics in these environments is far more complicated than the statement of the three golden rules.

Avatar and ethical rules to respect in virtual environment

Examining the literature on ethical codes that we should apply when working with avatars, we find that they are generally presented as a set of rules to follow (Berry, 2004). It is a set of guidelines formulated to help the researcher in his/her research process (Thomas, 1996). These definitions reflect the rigidity of the ethical codes design which is totally inadequate with the flexible nature that characterizes the online sphere and even more the virtual reality. To consider in detail one of these guides, we refer to the code of the Association of Internet Research (AoIR, 2002) which is the official code that researchers should use for conducting their online research. The working committee in charge of this code on the online research ethics published its first report in October 2001. The first goal of this committee was to develop a list of values that the researchers on Internet should follow. It stresses the importance of giving special attention to the social sciences online research and research practice in general. The report highlights the complexity of the task on the following points:

• ensure anonymity and confidentiality of individuals (the individuals, the avatar but also the group);

• obtain consent;

• verify the identity of the subjects;

• determine the most appropriate ethical approach considering the global nature of the internet.

The report does not provide universal guidelines: it only reiterates the points that we find in most conventional research involving human subjects. All agree that individuals must be respected and protected as autonomous agents (Jankowski and Van Selm, 2008). We wonder if this code succeeded to differentiate itself from the single standard codes and if the flexibility we were talking about has been taken into account. That criticism aside, it is still a good initiative that stresses the importance to consider the differences between virtual and real research.

Let us now review some of the fundamental rules of any code of ethics and how they can be translated into the virtual worlds considering their impact on the avatar. A basic principle of any ethical code is to protect subjects from potential harm. Thus, the definition given to damage to others must be reviewed: has it the same definition in the real world than in virtual world? In a utilitarian approach, the interest of the majority involves care not to prejudice the interests of the majority of "others”. Another approach is related to freedom: what are the limits of personal freedom in order to protect that of the others? These definitions lead us to ask ourselves whether the definition of "others" is the same in virtual worlds and whether the potential damages are the same. The position of a researcher could be that, fully aware of the “culture” of virtual life, he/she respects the same general and implicit ethical codes of life in these worlds. In this case, ethics in virtual worlds only relies on the goodwill of the researcher. An ethical code should not allow a researcher to freely define what a “damage” is. More and more people are involved in the virtual worlds, sometimes with strong emotional attachment (Dos Santos, 2008). Thus, we agree with one interpretation that has been made of the golden rules of research in cyberspace by Allen (1996) which stated that in online research, “harm” could occur. As a consequence, even if communication between participants and researchers has changed, a respectful dialogue, i.e. a conversation between two equal interlocutors, must be established by the researcher. An ethical rule suggested by the Avatar Identity Research Center might seem to fill a gap of AoIR. This rule speaks of the respect for the culture of the participants in their Second Life. Here, the key point is that, more than a respectful conversation which is the rule in online research, a researcher would have to know his/her participants before harvesting the data. To achieve this, he/she must learn about the participants by reading their public profiles in Second Life, which provide useful information to understand the avatar (and the person behind it). For example, the mere mention of real life, beyond the issue of confidentiality that will be discussed later, could be considered as an offence for an “avatar-participant” who in his/her public profile does not disclose any information about it.. The avatar being a kind of mask, to refer to the person behind the mask could be regarded as more sensitive than the most taboo topics in real life.

The “'Avatar Identity Research Center" (Costello, 2006) provides other rules related to research in Second Life, such as respecting confidentiality of the demographic characteristics. For the same reasons applying to the previous rule, the researcher must respect the confidentiality of certain demographic data which can be seen as important in some research. As these same data may be needed in some research, the data collection must be preceded by steps which respect the choices of people who have avatars with characteristics totally different from the ones in real life. Some people choose to change those characteristics in order to hide behind a new identity. The researcher must respect this choice and should not destabilize the interviewee. When demographic information is needed for the purpose of the study, the researcher must be careful to specify it in the consent form. People who do not want to comply will quit the research.

The consent is one of the key elements of any research either in a metaverse or in real life. The consent becomes more important in our case as already discussed before: not only some initial conditions must be announced to the interviewees but also, in relation with the complexity of the dual identities of the real and virtual worlds, the interviewees should be aware of the identity investigated by the study. Knowing the information sought, they will evaluate their willingness to disclose it or not and thus participate or not. Even if we look only at the avatars, we must nevertheless consider the scope of consent. Are avatars the only ones to be affected by the consent or should one also obtain a broader consent including the moderators, or even the Linden Labs (Mc Kee and Porter 2007)?

Even after an extensive literature review and an inventory of the expectations expressed by the avatars, we recognize that the topic of each study could generate different ethical situations (e.g. McKee and Porter (2007), not only because of the various topics but also because of rigid ethical codes specific to regions of the world. We are three researchers from three different regions of the world namely, North America (Canada), Europe (France), Africa (Tunisia), and have noted that the definition of research ethics was different. The wide platform that seems offered by Second Life could shrink through ethics. For example, if a code of ethics for online research in Canada may seem somewhat less flexible than in Europe, European researchers would thus have more freedom of action. But would it be really at the price of the avatar dignity? The same question arises at the level of private research (like business consulting) versus academic research. Ethics in academia which is probably much stricter, can become an obstacle for the researcher, who, even if attempting to move towards ethically freer virtual worlds (like Second Life), realizes that this apparent freedom is relative. The greatest proof of this is this document that we, researchers, have had the idea to write. Thinking ethics has become instinctive and automatic for a Canadian researcher when it could be regarded as a simple formality for a researcher working in another part of the world.

Conclusion

Virtual worlds, or metaverses, have experienced an explosive growth in the last decade. Research in these virtual worlds (like Second Life), is also growing (El Kamel and Rigaux-Bricmont, 2009). Many social scientists, including marketing ones, begin to see the methodological opportunities offered by using these metaverses as research fields. Indeed, the virtual interface has many advantages over traditional or online researches. They can be classified according to whether they are related either to contents or to tools for interaction with the virtual environment. Although methodologically appealing, research in metaverse is a new subject and its characteristics are not yet covered by ethical standards (often dating from an era in which technology was much less advanced). We have shown that research in Second Life has three advantages in terms of the classical ethical principles: 1. it promotes the easy removal of participants if they wish to leave the study; 2. it minimizes barriers between vulnerable populations and, finally, 3. it reduces some inconveniences for the participants. Beside from these advantages, some aspects of online research remain sensitive to ethics and complicate the task of the researcher interested in this type of experimental designs. In this sense, we have shown that certain characteristics of metaverses like Second Life make it difficult to apply traditional ethical rules and require the creation of other rules or guidelines to ethical conduct.

Bibliografía/Referencias

•  ALLEN. C, 1996, “What's wrong with the Golden Rule: Conundrums of conducting ethical research in cyberspace”. The Information Society, 12(2), 175-187.

• AMDITIS A., KARASEITANIDIS I. & MANTZOURANIS I, 2008, “Virtual Reality Research in Europe: Towards Structuring the European Research Area”. In D. TALAB? and A. AMDITIS (Eds.) Product Engineering: Tools and Methods Based on Virtual Reality, Springer Science, Business Media, pp.3-20.

• Association of Internet Researchers (AoIR) Ethics Working Committee, 2002, Ethical Decision-making and Internet Research: Recommendations from the AOIR Ethics Working Committee. Recovered 28 January 2009 from www.aoir.org/reports/ethics.pdf

• ASCH, S.E, 1951, “Effects of group pressure upon the modification and distortion of judgement”. In H. GUETZKOW (Eds.) Groups, leadership and men, Pittsburgh, PA: Carnegie Press, pp.222-236.

• ASCH, S.E, 1956, “Studies of independence and conformity: A minority of one against a unanimous majority”. Psychological Monographs, 70(9), 70.

• BERRY, M.D, 2004, “Internet research: privacy, ethics and alienation: an open source approach”. Internet Research, 14(4), 323-332.

• BONGRAND, P.C, 1905, De l’expérimentation sur l’homme. Sa valeur scientifique et sa légitimité. Doctoral Thesis, defended on 27 January 1905 at the Faculty of Medicine and Pharmacy of Bordeaux

• BRUCKMAN, A, 2001, “Studying the amateur artist: a perspective on disguising data collected in human subjects research on the Internet”. Ethics and Information Technology, 4(3), 217-231.

• CASTRONOVA, Edward, 2005, Synthetic Worlds: The Business of Online Games, The University of Chicago Press, 332 pages.

• CLARK, D, 2004, “What If You Meet Face to Face? A Case Study in Virtual/Material Research Ethics”. In E.A. BUCHANAN (Eds.) Readings in Virtual Research Ethics, Edition: Illustrated, pp.246-261.

• COLBURN A., COHEN M. F. & DRUCKER, S, 2000. “The role of eye gaze in avatar mediated conversational interfaces”. Technical Report MSR-TR-2000-81. Microsoft Research. Recovered, 28 October 2008 from http://ftp.research.microsoft.com/pub/tr/tr-2000-81.pdf.

• EL KAMEL L. et RIGAUX-BRICMONT B, 2009, “Une seconde vie aux méthodes traditionnelles de la recherche”, Université Laval, Document de travail, 15 pages.

• COSTELLO, C.Y, 2006, Ethical Rules for Conducting Research in Second Life. Département de Sociologie, Université du Wisconsin-Milwaukee.

• Commission nationale pour la Protection des sujets humains dans le cadre de la recherche biomédicale et béhavioriste, 1978, Rapport Belmont: Principes éthiques et directives concernant la protection des sujets humains. Recovered 23 December, 2008 from http://www.fhi.org/training/fr/RETC/pdf_files/FrenchBelmont.pdf

• CRSH, 2005, Éthique de la recherche avec des êtres humains. Recovered 28 January 2009 from www.pre.ethics.gc.ca/francais/pdf/TCPS%20octobre%202005_F.pdf

• CSIKSZENTMIHALYI M, 2002, Flow. London: Rider.

• DOS SANTOS, L, 2008. Le marketing de l’avatar sur Second Life: L’émergence de nouvelles questions sur les besoins. Essai de maîtrise, Département Marketing, Université Laval.

• FAGOT-LARGEAULT, A, 2000, Éthique de l’investigation scientifique sur l’être humain. Text from the 20ième Conference de l’Université de tous les savoirs given on 20 January 2000.

• EL KAMEL L., 2009, For a Better Exploration of Metaverses as Consumer Experiences, in N.T. Wood and M.R. Solomon, editors, Virtual Social Identity and Consumer Behavior, M.E. Sharpe Inc., New York.

• EL KAMEL L. and RIGAUX-BRICMONT B., 2009, Online Qualitative Research and Metaverses, Congress of CyberSociety, Topic D, Culture: Metaverses and Virtual Environments.

• FREIER, N.G, 2007, “Children distinguish conventional from moral violations in interactions with a personified agent”. ACM student research competition: Ext. Abstracts CHI ’07. ACM Press, (2195-2200).

• Gartner Group, 2007, Analysts Identify the Five Laws for Virtual Worlds. In Gartner Symposium, ITxpo 2007 Emerging Trends. Recovered 20, October, 2008 from www.gartner.com/it/page.jsp?id=503861

• GOFFMAN, E, 1959, The presentation of self in everyday life. Doubleday, Garden City, NY, USA.

• HELFER, J.P, 2003, Et si le marketing était éthique par définition. Papier de recherche. Université Paris I Panthéon Sorbonne. Recovered 10, Jannuary, 2008 from http://gregoriae.univ-paris1.fr/protect/cahier/2003-08.pdf

• HENDAOUI A., LIMAYEM M. & THOMPSON C.W, 2008, “3D Social Virtual Worlds. Research Issues and Challenges”, IEEE Internet Computing, (88-92).

• HOLBROOK, M.B, 1994, “Ethics in Consumer Research: An Overview and Prospectus”. In Advances in Consumer Research, 21, (566-571).

• JANKOWSKI, N & VAN SELM, M., 2008, Research Ethics in a Virtual World: Some Guidelines and Illustrations, Department of Communication, University of Nijmegen.

• JARVIS, J. 2009, “What Would Google Do?”,Haper Collins Publishers, 1^st, USA.

• JOINSON, A.N, 1998, “Causes and Effects of Disinhibition on the Internet”. In J. Gackenbach (Eds.) The Psychology of the Internet (pp.43–60), New York: Academic Press.

• LOEFFLER, T.A, 2004, “A Photo Elicitation Study of the Meanings of Outdoor Adventure Experiences”. Journal of Leisure Research, 36(4), 536-556.

• LUDLOW Peter and Wallace MARK, 2007, The Second Life Herald: The Virtual Tabloid that Witnessed the Dawn of the Metaverse, Massachusetts: MIT Press, Cambridge, 295 pages.

• MANU ALEXANDER, 2007, “The Imagination Challenge: Strategic Foresight and Innovation in the Global Economy”, Chapter 13, Duality in Dataspace, New Riders (203-219).

• MATHWICK, C. & RIGDON, E, 2004, “Play, Flow and the Online Search Experience”. Journal of Consumer Research, 25(dec.), 281-240.

• MCARTHUR, V, 2008, “Real Ethics in a Virtual World”. CHI 2008, ACM, Florence, Italy (3315-3320).

• Mc Kee, M. & Porter J.E, 2007, “Playing a Good Game: Ethical and Methodological Issues in Researching MMOGs” Recovered 28, Jannuary from www.wide.msu.edu

• MILGRAM, S, 1963, “Behavioral study of obedience”. Journal of Abnormal and Social Psychology, 67(4), 371-378.

• OCDE, 2008, La mesure du contenu crée par l’utilisateur: implications pour les enquêtes sur «l’accès aux TIC et leur utilisation par les ménages et les particuliers». Recovered 28 January 2008 from www.oecd.org/dataoecd/47/3/40008185.pdf

• Official Second Life Blog, 2009, Second Life Residents logged nearly 400 million hours in 2008, growing 61% over 2007. Recovered 28 January 2009 from http://blog.secondlife.com/2009/01/15/q42008/#more-3487

• Second Life, The Official Guide, 2007, Michael Rymaszewski, Wagner James Au, Mark Wallace, Catherine Winters, Cory Ondrejka, Benjamin Batstone-Cunningham, and Second Life Residents from around the world (Foreword by Philip Rosedale) , 342 pages.

• STERNBERG, J, 1998, “It’s All in the Timing: Synchronous Versus Asynchronous Computer-Mediated Communication”. Paper presented at 3rd Annual Conference of the New Jersey Communication Association Montclair, New Jersey.

• SOLLIE, P, 2007, “Ethics, technology developmentand uncertainty: an outline for any future ethics of technology”. Journal of Information,Communication & Ethics in Society, Vol. 5 No. 4, 293-306.

• STEPHENSON, N, 1992, Snowcrash. NY,USA: Bantam Books,

• Stewart, K. & Williams, M, 2005, “Researching online populations: the use of online focus groups for social research”. Qualitative Research, 5(4), 395-416.

• THOMAS J. (1996), “When cyber-research goes awry: the ethics of the Rimm ‘cyberporn’ study”, Information Society, 12(2), 189-97.

• TYBOUT, A. & ZALTMAN, G, 1974, “Ethics in Marketing Research: Their Pratical Relevance”. Journal of Marketing, 11(4), 357-368.

• USC, 2008, Annual Internet Survey by the Center for the Digital Future Finds Shifting Trends Among Adults About the Benefits and Consequences of Children Going Online. Recovered 28 January 2008 from www.digitalcenter.org/pdf/2008-Digital-Future-Report-Final-Release.pdf.

• WITMER, D.F., COLMAN R.W. et KATZMAN, S.L, 1999, “From Paper-and-Pencil to Screen-and-Keyboard”. In Jones (Eds.) Doing Internet Research (pp.145-161), London: Sage.

Recoñecemento – SenObraDerivada (CC-by-nd 3.0 ou calquera outra posterior)