Abstract
The Science and Technology Agency (STA) has been performing technology forecasts using the Delphi method approximately every five years since 1971. The survey will be continued as a basic tool for planning in Japanese R&D policy and, therefore, it is important to evaluate and improve its reliability. This is important also for international harmonization of technology forecast methods, considering that Germany and Britain have recently started similar long-term forecasts. The present work analyzes the reliability of results obtained in the newest 5th survey with particular emphasis on appropriateness of the respondents who included specialists as well as people of less expertise in relevant disciplines. It is indicated that respondent groups consisting only of experts will improve the reliability of the survey.
