Instrumentation for Experimental Economics Laboratory

在美国国家科学基金会的支持下，弗农·史密斯博士将为亚利桑那大学经济科学实验室升级和扩建一台便携式高分辨率投影系统，购买20台90 MHz奔腾‘IBM类型兼容’微型计算机、27台16MB内存升级现有计算机和一个便携式高分辨率投影系统。史密斯博士是实验经济学的领军人物，他和他的合作者致力于开发实验情景，在这种情景中，人类受试者做出具有真正潜在经济回报的选择。计算机化使设计各种实时市场机构成为可能，在实验室研究它们的表现，从而洞察市场如何运作和人类如何反应。从大约十年前开始，实验室实验开始探索“智能”计算机辅助市场的表现。在这些实验市场中，去中心化代理的消息(出价)使用各种优化算法来处理，这些算法应用于批量出价或实时到达的出价。这极大地扩大了对硬件容量和软件复杂性的要求。随着这些实验市场在技术上变得更加复杂，对系统设计和支持的要求也越来越高，它们也对人类受试者表达基于判断和计算的理性回应信息的能力提出了更高的要求。因此，下一代试验性市场将需要在每个站点为该站点的决策者(S)提供“专家系统”计算机支持。本奖项中提供的计算机将使这种进步成为可能。除了增加对构成人类经济行为基础的原理的理解外，实验经济学还具有许多实用功能。它帮助监管机构和政府官员设计系统，以分配无线电波空间或航空公司登机口等稀缺资源，并设计有效的竞标程序。在亚利桑那州，开发的课程还将纳入当地和世界范围内的经济学教学。

With National Science Foundation support Dr. Vernon Smith will purchase 20 90 mHz Pentium `IBM type compatible` micro computers, 27 16 MB RAM upgrades for existing machines and a portable high resolution projection system to upgrade and expand the University of Arizona Economic Science Laboratory. Dr Smith is a leader in experimental economics and he and his collaborators have worked to develop experimental situations in which human subjects make choices which have real potential economic rewards. Computerization has made it feasible to design various real time market institutions, to study their performance in the laboratory and thus gain insight into how markets function and how humans respond. Beginning about a decade ago laboratory experiments started to explore the performance of `smart` computer assisted markets. In these experimental markets the messages (bids) of decentralized agents are processed using various optimization algorithms applied to batches of bids or to bids arriving in real time. This has greatly expanded the demands on hardware capacity and software sophistication. As these experimental markets have become technologically more complex, and demanding of system design and support, they have also become more demanding on the ability of human subjects to articulate rational message responded based on judgment and calculation. Consequently the next generation of experimental markets will require `expert system` computer support at each station for the decision maker(s) at that station. The computers provided in this award will make such an advance possible. In addition to increasing understanding of principles which underlie human economic behavior, experimental economics serves a number of practical functions. It helps regulators and government officials to design systems to allocate scarce resources such as radio wave space or airline gates and to design effective bidding processes. At Arizona the programs developed will also be incorporated into the teaching of economics on both a local and worldwide basis.