Website Accessiblity

Research Topics

Human Computer Interaction

The field of human-computer interaction is a confluence of many disciplines: cognitive psychology, artificial intelligence, software engineering, computer graphics, and human factors. The overall goal of the field is to improve the design of computer interfaces in order to facilitate the learning and usage of computer systems. The purpose of current research is to develop cognitive models of users performing specific tasks with computer systems for office automation and decision support. These cognitive models represent the knowledge the user has of both the task and the operating procedures for performing the task with the computer system. The models can be seen as theories for making quantitative predictions of the ease of use and the ease of learning for different system designs. In practice, these predictions are used to improve interface procedures and to enhance the design of user assistance and training.

Back to top

Software Productivity

Applications of computer and communications technologies are facilitated through software. The development of software is notoriously labor-intensive, and this often is a limitation for new opportunities. It is known that comparable system products may be delivered at different levels of efficiency and cost. But it is not well understood how to manage software development activities more productively. How should software development productivity be measured? Besides labor in general, what are the key environmental factors that drive software development costs? One aspect of this research concerns an empirical study of system-development project history in large organizations. Several software metrics are employed to monitor system products and to compare input resource consumption. Research concerning software maintenance management has also been undertaken.

Back to top

Evaluation of Information Systems

A current, common procedure for the evaluation of an information system employs user surveys. Surrogate measures, such as user satisfaction and the level of system usage, are used to determine a system's effectiveness. But the many problems associated with the administration and use of user evaluations are well documented. How can the evaluation procedure be improved? Can a theoretical framework upon which the evaluation process might be based be developed? How does the information system affect the various measures of effectiveness of an organization; e.g., productivity, responsiveness to customers? Current work is being carried out to develop such a framework by expanding on methods and results from the fields of microeconomics, information economics, and statistical decision theory. Of interest is modeling and measuring the impact of the information system on the performance of an organization or a worker. Can the relevant attributes of information and an information system be analytically defined? How do these attributes relate to the value or usage of an information system?

Back to top

Corporate Telecommunication Networks

Within the last five years, major corporations have dramatically accelerated the installation of user-controlled telecommunications facilities, replacing the purchase of telecommunications services from the carriers. Continuation of this trend could alter the roles of carriers and customers, lead to major shifts in the types of facilities and services demanded from carriers and equipment suppliers, and transform the role and function of the corporate telecommunications department. Research in this area is exploring the underlying cost structure of user owned facilities versus carrier services, and the impacts both on the organization and on carriers of the switch to user-controlled facilities.

Back to top

Business Value of Information Technologies

A major question confronting the management in U.S. industry today is whether or not the anticipated economic benefits of investment in Information Technologies (IT) are being realized. The results from reported prior research on this question are at best mixed. Current research proposes a theory and methodology for determining the economic impacts or "business value" of IT. The research is planned in two phases. The first phase involves a statistical analysis of industry data acquired by the investigators to test several conjectures which may explain prior discrepancies. Based on these results, the second phase develops a microeconomic model of IT investment decisions for "information sector" industries. This model would then be pilot tested employing a new data set to validate-or disaffirm-the theoretical results.

Back to top

Software Production as Scientific Discovery

Despite 50 years of progress, the software industry is unable to meet the demands of the information age society. We are exploring a new paradigm: viewing software production as a special case of scientific discovery. In this paradigm, programmers develop rules (i.e., programs) to solve sets of phenomena, test these rules by running experiments, and search for new rule representations to innovate. Our current focus is on Object-Oriented Design and Object-Oriented Programming. In order to test this new paradigm, we have built a market for software components (i.e., objects) in which programmers sell and buy rules (i.e., objects), assemble and test these rules into programs, and then evaluate the effectiveness of their programs. A key aspect of this software market is how buyers and sellers decompose and represent the functionality of their software components.

Back to top

Computer Support for Real-Time Dynamic Decision Making

For the most part, research efforts in the last two decades have been directed at understanding and supporting decision making activities in static decision tasks. This research focuses on how to support individuals and groups in real-time environments with interdependent decisions. Current efforts trace decision making in organizations in order to understand how norms, information flows, organizational values and assumptions constrain decision making. These studies are complemented in the laboratory by building computer simulations and testing the impact of computer support on overcoming cognitive limitations. In the practical side, these studies generate new methodologies and tools for designing and implementing Information Technology in complex decision making situations.

Back to top

Connecting Households to the Information Highway

Although government and many businesses are preparing to offer advanced electronic services in the near future, no study of the potential benefits and impact on families has ever been done. The HomeNet project at CMU will answer central questions about how families at home will be able to use advanced communication and information services. It will include a panel of families using advanced information services, including access to the Internet. HomeNet is unusual in that we will offer both electronic interaction and communication in a residential community and the use of particular services by these families. HomeNet will provide a platform for addressing issues in four domains. What makes electronic services valuable to families? What relationship would electronic services to households have with education? What impact will the services have in the household and the larger community? How can electronic services evolve?

Back to top

Electronic Data Interchange and Interorganizational Systems

The growth of interorganizational systems (IOS) in recent years has been impressive. For example, the computer to computer exchange of data between buyers and suppliers using standard format has been increasing sharply due to the Electronic Data Interchange (EDI) technology. However, many research issues arising from these systems are yet to be examined. For example, would the growth of such systems stall due to negative externalities thus compelling the system initiator to provide subsidies to the reluctant trading partners? Would IOS and EDI lead to interdependent but unequal benefits to buyers and suppliers? Can EDI improve information attributes such as accuracy and timeliness and thus lead to higher performance of trading partners? What are the benefits of increasing vertical information integration using EDI technology in JIT environments? Several research projects examining the theoretical and empirical issues raised by IOS and EDI have made significant progress in the last few years.

Back to top

Footer Navigation