科学計量学の手法(<特集>研究・技術計画/政策科学のメソドロジー)

抄録

The article reviews scientometrics as a methodology for policy science. Science and technology policy necessitates measurement and evaluation of the current scientific activities in order to develop future steps. Scientometrics, which studies methods, objects and effectiveness of measuring scientific activities, should therefore be an important constituent of policy science. The term "scientometrics" was coined after "psychometrics" and "sociometrics", which share quantitative view of the subject of study. The present work first shows how scientometrics differs from bibliometrics or library and information science to indicate the need of a theory on what aspect of scientific activity is measured quantitatively. It then presents a historical overview of scientometrics in four phases: pre-history, establishment of the principle, development and institutionalization of indicators, and application to evaluation. The article further reviews practical methods including analysis of the number of reports, citation and co-citation analysis, term and co-term analysis, and forming and combination of indicators. Discussion on the possibility of and potential issues on quantification concludes the article.