Tridas Mukhopadhyay - US grants

The objective of this project is to develop a new psychological model of expert software effort estimation based on analogy and case-based reasoning. The research activities include psychological studies, computer modeling and validation, and simulation experiments to formulate an accurate model of analogical reasoning in the domain of effort estimation. The results of this research will provide major insights into how expert effort estimators perform their task and how to improve the effort estimation process. The long term goal is to weave together the symbolic methods of the expert with analytical methods in order to provide more understandable and more accurate estimates of project effort.

A major question confronting management in U.S. industry is whether the anticipated economic benefits of investments in information technologies are being realized. the results from prior reported empirical research on this question are at best mixed. The researchers consider this problem to be a measurement related issue, and they develop a theory and methodology for determining the economic impacts of "business value" of information technology. The focus of the research is on the firm and its profit centers as the unit of analysis. The research is planned in three phases. The first phase involves a statistical analysis of manufacturing industry data acquired by the investigators to test several conjectures on prior discrepancies and estimate a theoretical model. Based on these results, the model is refined in phase two and further tested (validated) with two new data sets from banking and insurance industry firms. The third phase synthesizes management implications of the research for information technology investment decisions.

This research examines how household computing and the
Internet are changing the social and psychological well being of the
American family, and compares the influence of these new technologies
on households with that of older ones--telephones and
televisions. The research objectives are (1) to determine whether the
early results showing that use of the Internet decreases psychological
and social well being generalizes across samples, Internet services, and
time, (2) to identify mechanism for these effects, and (3) to compare
the impact of different patterns of household information technology
use. Data will come from a national three-wave, panel survey, allowing one
to examine how the amount of television, telephone, computing, and
Internet people use at one time predict later changes in their social
involvement and psychological well being at a subsequent time. The
survey will be supplemented with a quasi-experiment, in which a subsample
of respondents who have computers but no Internet access will be
randomly assigned free Internet access instead of a cash payments to
participate in the survey. Results will have implications both for
emerging policy debate about universal access to the Internet and for
the design of socially responsible Internet services.

This research provides a comprehensive, integrated theoretical framework to examine the relationship between information technology substitutes and complements and work group performance. The methodological approach uses an organizational learning curve framework in which to examine the effects of incremental investments in information technology (IT) on work group productivity, mediated by the effects of IT investment upon the learning curves experienced by the work groups. The research uses a rigorous mathematical model that is grounded in practice and theory, and complementary to existing organizational behavior and organizational learning theories. It is hypothesized that IT that substitutes for human resources results in lower organizational learning and higher knowledge depreciation, while IT that complements human resources results in higher organizational learning and lower knowledge depreciation. The model instantiates a multi-dimensional view of group performance by incorporating in the learning curve dependent variable not only productivity measures, as in traditional learning curve models, but also measures of cost and quality. The investigation will test the integrated model empirically in the work setting of a major financial institution. This research potentially has very broad implications for the consequences of different IT investments on organizations -- public, private, and government -- their productivity, quality, and costs.