| The ability to anticipate future capabilities of technology products has broad implications for organizations. Technological forecasting allows management to improve resource allocation, make better staffing decisions, and more confidently plan facilities and capital expenditures. Technology forecasting may also identify potential new markets and opportunities, such as finding ways to exploit current technology beyond its originally intended purposes.; Modern technology forecasters use an array of forecasting methods to predict the future performance of a technology, such as time-series analysis, regression, stochastic methods, and simulation. These quantitative methods rely on the assumption that past behavior will continue. Shortcomings include their lack of emphasis on the best technology available and the fact that they do not effectively address the dynamic nature of ever changing trade-off surfaces. This research proposes a new method to address the shortcomings of common forecasting techniques by extending a well-established management science methodology known as data envelopment analysis (DEA). This new method is referred to as Technology Forecasting with Data Envelopment Analysis (TFDEA). Three case studies are examined to determine the method's validity.; The first case study is that of relational database system performance based upon industry benchmarks obtained from the Transaction Processing Performance Council (TPC). The results reveal that TFDEA provides a more accurate picture of the state of the art than basic regression. The second case study expands Moore's law to six dimensions, resulting in a more comprehensive assessment of microprocessor technology. The final case study re-examines hard disk drive data for the years 1994--1999 in order to evaluate the technological progress of multiple technological approaches presented in Christensen's The Innovator's Dilemma.; Major contributions include both a new technology forecasting technique and an important extension of the temporal DEA methodology, which together offer a new and more comprehensive method for evaluating and forecasting technology. |
