| Budget Amount *help |
¥2,400,000 (Direct Cost: ¥2,400,000)
Fiscal Year 1998: ¥700,000 (Direct Cost: ¥700,000)
Fiscal Year 1997: ¥600,000 (Direct Cost: ¥600,000)
Fiscal Year 1996: ¥1,100,000 (Direct Cost: ¥1,100,000)
|
|---|
| Research Abstract |
The goal of this project was running experiments using subjects of several cultures in the same session. We first needed a computerized experimental laboratory capable of -developing and managing multilingual experiments over the global Internet. Assembling the system took most of the first year. We-first used an ordinary web server and a CGI program scripted in Pen. Dynamic web pages provided subjects with information about the decisions they needed to make, and included forms to return the decision to the server, Successful for very simple games, such as the Prisoner's Dilemma or the Ultimatum Game, it is too clumsy for continuous time experiments such as auctions. For multistage games such as Rosenthal's ''Centipede Game'' it significantly slows the experiment. Recoding in C or C++ showed no significant speed benefit. Future efforts will use Java as a cross-platform progamming environment which facilitates real-time updates (eg, in market experiments) using the ubiquitous web browse
… More
r as the subject workstation, and on a dedicated ''networked experiment protocol'' for more advanced environments.
Some simple experiments quickly showed that the Microsoft Windows environment is insufficient. First, although the Microsoft ''Visual'' series of compilers Are easy to use to develop user interfaces, ' it is a very restrictive environment for developing network applications. Second, multilingual development is difficult in the Windows environment ; using more than English plus another language simultaneously requires rebooting the computer. For this reason we are developing our next version using Linux, an open source operating system. Linux provides the following features : excellent networking, including emulations of the primary Microsoft network protocols ; multilingual editors such as XEmacs and yudit, and other internationalization support ; the GNU gettext library, which makes possible parallel development of a given experimental' protocol in several languages ; and revision control and achiving systems such as CVS. Less
|
