6th conference on New Directions in Management Accounting, Brussels, 15-17 Dec 2008
Plenary presentation on
From the Laboratory to the Virtual World
Robert Bloomfield (Cornell University, USA)
Abstract
Virtual World technology provides a variety of business and research opportunities. I will begin my talk by describing how enterprises—from IBM, Samsung and SAP to FASB and NASA—are using virtual worlds to collaborate across long distances with employees and partners in strategic alliances. Virtual worlds allow for such communication in a manner far superior to more familiar technologies (such as conference calls) because of their ability to give people a very personal presence in a seemingly real place; allow for a combination of text and voice communication; and the ability to support 3D prototyping and other content creation. Research is needed to examine the efficacy of such interactions; the impact of virtual world technology on relationships among employees and partners; and the optimal governance of increasingly virtual organizations.
I will then discuss how experimental researchers in managerial accounting can use virtual worlds to free themselves from the confines of the laboratory. Virtual worlds provide three key advantages to the researcher: people, time and tools. The typical laboratory experiment can accommodate only a couple dozen interacting participants over the course of a couple of hours, using rather crude software tools developed by academics on limited budgets. In contract, virtual worlds allow researchers to attract large numbers of participants who might interact over weeks or months in software platforms developed by professionals with budgets approaching $100 million. People, time and tools allow researchers to examine institutions more rich and complex than possible in the laboratory, creating opportunities to examine new research questions, and to test the robustness of laboratory effects in more complex settings.
Researchers using virtual worlds for experiments will need to address two types of challenges. First is the challenge imposed by the complexity that virtual worlds allow: researchers must make sure that they can administer complex experiments reliably, and can draw clear inferences from settings in which much behavior is endogenous (resulting in the use of many measured independent variables). Second is the challenge inherent in any online experiment: a lack of control of participant behavior. Researchers should conduct experiments in virtual worlds only if the benefits of access to people, time and tools outweigh the costs of administration, inferential difficulty and lack of control over participants. I close by discussing some examples in which the move from the laboratory to the virtual world is (or is not) appropriate.