Patterns of Research Productivity in the Virtual Worlds Literature: A Bibliometric Approach

Since the beginning of the 21st century, the internet has provided a much more dynamic platform, becoming more visual and allowing users to interact with each other in communities and share information in real time. The new Internet is a social media, and online communities are taking on a new dimension, quite literally, in the form of immersive 3D virtual worlds, such as such as Active Worlds, OSGrid, and Second Life. Surprisingly, 80% of active internet users have a registered account in virtual worlds which are playing an increasingly important role in the lives of many adults, teens, and children. Indeed, virtual worlds are believed to have implications that go beyond how we play, to also include how we buy, work, and learn. In recent years there has been increased behavioral research in virtual reality and virtual worlds. As a result, the subject of virtual world has become a major research area in recent years. To shed light on virtual worlds’ trends, and contributions, a historical review and Bibliometric analysis are included in this research. The Bibliometric analytical technique was used to examine this topic in SSCI journals from 1991 through 2011. Also, the study investigated whether author productivity conformed to Lotka’s Law, to understand author distributions within the field of virtual worlds and to elucidate the distributions of the core authors.


Introduction
Since the beginning of the 21st century, the Internet has provided a much more dynamic platform, becoming more visual and allowing users to interact with each other in communities and share information in real time.The new Internet is a social medium, and online communities are taking on a new dimension, quite literally, in the form of immersive three-dimensional (3D) virtual worlds.A virtual world, sometimes also called digital worlds or Metavarse, is a computer-based simulated environment, usually modeled after the real world, accessed through an online interface, and inhabited by users in the form of avatars-a representative of a real person in virtual world (Kaplan and Haenlein, 2009).
The high growth of various virtual worlds globally that has occurred in recent years has prompted a number of Fortune 2,000 companies to either enter the virtual space, or at least monitor the development and potential of avatar-based business, such as Apple, IBM, Nike, GM and Disney, have begun to extend their businesses to virtual worlds.Even the American Cancer Society has used Second Life to promote its annual event.The use of virtual worlds is beginning to grow rapidly far beyond people's expectations (Hua and Haughton, 2009).
The current virtual worlds have their roots in different text-based environments, for example, discussion forums and multi-user dungeons (MUDs), which emerged as early as in the late 1970s and gained wide success in the 1990s when Internet became commonplace.One of the main advantages of virtual worlds is overcoming the limitations related to geographical distances (the need to commute).During the last few years, 3D virtual worlds have finally gained global popularity.The most advanced environments provide their users with detailed 3D graphics, animation, different communication methods including voice communication, features for personalization and building new objects, and a massive number of places and objects created by others already available (Partala, 2011).
Virtual worlds include everything from massive multiplayer online role-playing games (MMORPGs), such as World of Warcraft (WoW), to virtual realities such as Second Life.The business world has assumed that virtual worlds can be leveraged to provide access to consumers and consumer data (Spaulding, 2010).In recognition of the growing importance of avatars as 3D representations of people and their alter egos in virtual worlds, Google has a project underway to develop "universal" avatars that can move between virtual worlds.Some authors even suggest that the 3D Internet will become as important to companies in five years as the Web is now (Messinger, Stroulia, and Lyons, 2008).
In the last decade, the digital revolution affected various social-economic areas.Among the many venues of research, nowadays the attention is gathered by two distinct, yet interrelated, innovative threads: the so-called Web 2.0 and the rise of virtual worlds.As both phenomena are recent and highly dynamic they have not been clearly understood and explained yet.Virtual worlds have captured the interest of many scholars and practitioners, in particular because the entrance of many real world companies, attracted by the potentialities of virtual worlds, has been interpreted as the signal for the discovery of new web-based business models, in the same vein the boom of Internet did in the past (Cagnina and Poian, 2009).
In this paper, we are going to direct our efforts on the rise of virtual worlds, in order to understand their features.Therefore, we explored virtual world's research trends and forecasts by means of bibliometric from 1991 through 2011 to elucidate the virtual worlds' trends in adopting enterprises, contributions that virtual worlds are making, and forecasts for virtual worlds' growth.

What is "virtual worlds"?
Virtual Worlds are virtual places that enables users to communicate, cooperate, and collaborate, as in the real world (Hindmarsh, Heath, and Fraser 2006), which are considered a sub-set of the vaster field of Web 2.0 themes (Cagnina and Poian, 2009).Virtual worlds are understood as immersive, 3D, multimedia, multi-person simulation environments, where each participant adopts an alter ego and interacts with the world in real time.World activity persists even if a player is off-line (Wagner and Ip, 2009).
According to Harrison, Haruvy, and Rutström (2011), the virtual worlds simply mean something that is like something else, but without some of the properties of that which it is like.Although that sounds vague, some examples illustrate why one would want a general definition.In the context of computers, the word virtual just means computer simulated.Thus, a virtual environment implies that something inherent to the physical environment in which a user is communicating, socializing or gathering information is computer generated or mediated.Virtual worlds are perhaps the most extensive form of virtual environments, involving visual simulations that allow interactions between people in real time.Eisenbeiss et al. (2012) define virtual worlds as an unstructured social and technological environment that possesses three central characteristics: (1) It is embedded in a three-dimensional, visually sophisticated digital space; (2) it comprises an aggregation of people who are graphically represented by avatars, and movements of these avatars within the digital space are rendered simultaneously to all other constituents in the virtual vicinity in a three-dimensional visualization that enables real-time interaction; and (3) users of the virtual worlds engage in different exchange processes, whether social (mutual dissemination of thoughts and opinions), material (trading virtual material objects), or monetary (transfer of virtual currency).Every constituent engages in some but not necessarily all exchange processes.This definition distinguishes virtual worlds from related virtual venues, such as multiplayer online games or virtual communities.

What are the features of virtual worlds?
Generally, virtual worlds is "an electronic environment that visually mimics complex physical spaces, where people can interact with each other and with virtual objects, and where people are represented by animated characters" (Bainbridge, 2007).Open or unstructured virtual worlds represent a blending of the elements of immersive 3D gaming environments, developed in the gaming industry over the last 25 years, together with elements of online social networking.This conclusion can be seen by tracing the development of electronic gaming since the 1970s, including (a) arcade games, (b) console games, (c) LAN games with more players, (d) games with Internet connectivity, (e) unstructured games with many players, (f) massive games with user-generated content, and (g) immersive 3D worlds with designer-provided objectives.Open virtual worlds combine the last three items with elements of web-based social networking.Open virtual worlds, thus, consist of massively multiplayer gaming platforms with unstructured objectives, user-generated content, immersive 3D virtual reality shared environments, and social networking elements used between people through their avatars (Messinger, Stroulia, and Lyons, 2008).Cagnina and Poian (2009) outline that the founding characteristics of virtual worlds are: 1.
Interactivity: it exists on one computer but can be accessed remotely (i.e. by an Internet connection) and simultaneously by a large number of people, with the command inputs of one person affecting the command results of other people.

2.
Physicality: people access the program through an interface that simulates a first person physical environment on their computer screen; the environment is generally ruled by the natural laws of Earth and is characterized by scarcity of resources.

3.
Persistence: the program continues to run whether anyone is using it or not; it remembers the location of people and things, as well as the ownership of objects.
Furthermore, there are three-dimensional spaces perceived by users in virtual world, which provide them with an interactive experience with virtual objects, as well as with other virtual worlds' denizens (Saunders et al., 2011): 1. Perceptual space is defined as "that can be seen or sensed at one place and at one time," which is created by manipulating objects, interacting socially, and otherwise experiencing the virtual world through the senses.

2.
Cognitive space is the large-scale space beyond the sensory horizon about which information must be mentally organized, stored, and recalled in which users have experienced the virtual world through their senses.

3.
Universal space is defined as an"image of the environment" or a relatively stable system based on experiences with things and others.
In comparison to other social media, virtual worlds have three characteristics that differentiate them from other applications.First, virtual worlds allow users to interact with others in real time.Second, virtual worlds allow users to create fully customized virtual self-presentations in the form of avatars.Finally, while content communities, blogs, and collaborative sites are two-dimensional (i.e., focused on content sharing), avatars within virtual worlds have the possibility of exploring their virtual environment in three dimensions (Kaplan and Haenlein, 2009).

What are the studies of virtual worlds?
The literatures of virtual worlds can be classified as the following categories: 1.
Studies that investigate the link between individuals and their avatars in VWs as well as the impact of VWs on virtual identity building (e.g., Bélisle and Bodur, 2010;Parmentier and Rolland, 2009); 2.
Studies that analyze the experiences in the VW and consumer purchase behavior in the real world (e.g., Animesh et al., 2011, Gabisch, 2010;Haenlein and Kaplan 2009).

Methodology
All documents used in this study were accessed from the database of the Social Science Citation Index (SSCI), obtained by subscription from the ISI, Web of Science, Philadelphia, PA, USA.In this study, we only discuss the papers published in the period beginning 1991 because there were less data regarding virtual worlds prior to that year.
To shed light on virtual worlds' trends and contributions, a bibliometric analysis was conducted in this research.For the bibliometric analysis, the SSCI was systematically searched for virtual worlds-related materials published from 1991 through 2011.Selected documents included ''virtual world*'' in the topic.Analyzed parameters included authorship, patterns of international collaboration, journal, language, document type, research address, number of times cited, and reprint author's address.
The analyses included whether the growth and distribution of literature and author productivity are in compliance with Lotka's Law, with the purpose of understanding the growth and types of virtual worlds literature as well as the major research institutions.
Lotka's law' describes the frequency of publication by authors in a given subject field, it stated that "... the number (of authors) making n contributions is about 1/ n 2 of those making one: and the proportion of all contributors, that make a single contribution, is about 60 percent."This means that out of all the authors in a given subject, about 60% publish only one article, 15% (1/2 2 times .60)publish two articles, 7% (1/3 2 times .60)publish three articles, and so on.According to Lotka, only 6% authors in a subject field produce more than ten articles.
Lotka's law is often called inverse square law indicating that there is an inverse relation between the number of publications and the number of authors producing these publications (Zabed Ahmed and Anisur Rahman, 2009).

Publication statistics
The key word ''virtual world*" was used to search SSCI entries from 1991 to 2011 (updated December 31, 2011).
The virtual worlds SSCI article distribution status was used for trend analysis.

Distribution by publication year, document type and language
Table 3 displays virtual worlds-related publications by year, document type, and language.Figure 1 shows the number and percentage of annual publications output.In the SSCI, articles comprised the majority of published virtual worlds document types (Table 3).As for distribution by language, we see from Table 3 that the majority language of virtual worlds research is done in English.One interesting finding is that the publication had a drop in 2002 and 2004 respectively, and there has been a re-increase in virtual worlds research since 2005; it is clear that virtual world is becoming ever more important.

Distribution by source title
Table 4 shows that ''Cyberpsychology Behavior", "Computer Education,'' ''Presence Teleoperators and Virtual Environments,'' "Computer in Human Behaviors," and "British Journal of Education Technology," are the journals with the most publications on virtual worlds.In total, there are 768 journals had published 1,731 virtual worlds literatures.Through the 768 journals, 544 journals only published one literature among those journals.

Distribution by authors
From the retrieved 1,731 literature entries between 1991 and 2011, we calculated the number of authors using the equality method; that is, each author in a multiple-author literature is regarded to have an equal contribution.There are a total of 3,557 authors with an average of 1.17 publications per person.The largest amount of literature published by one person is 25 publications (1 person).The number (4 people) of authors who produced more than 10 articles is quite small (only 3%).and there are 3,201 authors who only published one paper, which accounts for 89.99% of the total publications.The total published literature deduced from authors using the equality method should be 4,170; however, the actual number of literature entries is 1,731.It is estimated that each piece of literature is written by an average of 2.41 authors and that the size of each research team is two to three people.The study of authors of virtual worlds literature between 1991 and 2011 can be explored using Lotka's Law.To validate whether Lotka's Law is applicable for virtual wolrds' literature, the slope value n and the constant value c should be calculated first, and a K-S test should then be performed to determine whether the distribution is in compliance with Lotka's Law (Bailón-Moreno et al., 2005).From the basic data, 89.99% of the authors have only one publication on virtual worlds in the last two decades studied; this result is different from the c value of 60.79% that appears in the original Lotka's Law.Therefore, the n value and the c value are calculated using the method of least squares to further examine their compliance with the Lotka's Law.
The calculated results in Table 6 are introduced into the calculation formula and the slope value n can be obtained as follows: The calculated results in Table 6 are introduced into the calculation formula and the slope value n can be obtained as follows: After obtaining the slope, the n value is introduced into another formula, to obtain the c value: When n= -2.802 and c=0.80223449, then f(x) = 0.80223449/x 3.1015 can be deduced.For n and c values, the n value is about -2 and the c value is 0.8079 in the original Lotka's Law, which indicates that the distribution of authors in virtual worlds literatures does not completely match with the original Lotka's Law.Pao (1985) described a least-squares (LS) method for testing Lotka's law, and suggested the procedures for computing values of the exponent n and constant c and the subsequent Kolmogorov-Smirnov (K-S) goodness-of-fit test for conformity.Zabed Ahmed and Anisur Rahman (2009) found that the maximum likelihood (ML) is a good method and that there is a tendency for least-squares fits to overestimate the slope of the power law since the statistical fluctuations in the logarithms of the data are greater in the downward direction than in the upward one.
Therefore, the LS method is used to estimate the best-fitting value for the slope of a regression line which is the exponent n for Lotka's law.The slope is usually calculated excluding high productive authorship from the dataset.Since values of the slope change with different number of author data, several computations of n were made.The n value as 2.058 provides the best-fitting value for the dataset.The n value was also calculated by ML method using Latka program.The ß-value (the Lotka exponent) is 2.1569 for authorship data.
The K-S statistic was performed again to test the observed and estimated values of LS n = 1.9022.The maximum absolute difference Dmax is 0.845 which falls outside the critical value of 0.0423 at 5% significance level.However, Kolmogorov-Smirnov statistics for n = 2.058 found D value 0.0389 which is max within the critical value at the 5% significance level.The Lotka program for K-S statistics for ML distribution is 0.0257 which is also below the 5% critical value of significance and hence, both should be accepted as appropriate models for the dataset.

Table 1 .
Distribution of virtual worlds published in each year between 1991 and2011 Table 1 indicated that all 1,731 papers found in the search were analyzed.4.2.Distribution by country/territory and institution nameTable2shows the distribution of publications by country and territory.The US, England, Germany, Canada, Australia, Italy, Netherlands, Spain, China, and Taiwan were the top ten countries publishing virtual worlds literatures.Listing publications by institution name, Table2also shows that the University of Illinois, University of Central Florida, Indiana University, University of South California, and University of Washington are the top five virtual worlds research institutions.

Table 2 .
Distribution by country/territory and institution name

Table 3 .
Distribution by publication year, document type and language

Table 4 .
Distribution by source titleTable5shows that ''Psychology", "Computer Science,'' and ''Information Science and Library Science,'' "Education Research," and "Business Economics," were the five most frequently used key words appearing in virtual worlds publications.

Table 6 .
Analysis of authors and their productivity of virtual worlds literature