- 期刊
台灣社會科學量化資料再用之研究：2001-2015年
The Reuse of Quantitative Data in Social Sciences in Taiwan: 2001-2015
摘要
本文分析2001-2015年間，社會科學領域利用既有量化資料從事具 原創性研究之狀況，取TSSCI 2015年收錄之經濟、政治、社會、教育與心理共五學門57種期刊，辨識出1,484篇再用論文，佔整體實徵論文17.38%，經濟與政治學門比例最高，心理則非常稀少。全部被引資料筆數共2,990筆，亦多由經濟與政治論文使用，且兩大學門論文同時使用多筆資料為常態；社會與教育論文則以使用一筆資料居多。在資料來源與資料類型上，政府機關資料及業務數據為被用大宗，但不同領域在資料來源與類型上有明顯差異；個別研究與一次性資料再用程度都很低。本文建議社會科學的資料庋用服務應著力於蒐集並媒合公私部門的大規模、長期性數據資料，並建立跨政府與民間的服務機制，當可裨益資料的充分再用。
並列摘要
This study explored the reuse of existing quantitative data in original social sciences research in Taiwan. Using the 2015 TSSCI List as the basis of journal selection, this study focused on 57 journals listed under the subject divisions of economics, political science, sociology, education, and psychology. Journal issues published between 2001 and 2015 were manually scanned to identify data reuse papers. The characteristics of the reuse papers and the cited data were recorded for the subsequent subject division-level and subject discipline-level analyses. A total of 1,484 reuse papers were identified, which accounted for 17.38% of the total empirical study papers. Among the five subject divisions, economics and political science had the highest percentages of data reuse, while psychology had the lowest. Those reuse papers together cited 2,990 datasets. Most of the datasets were used in economics and political science papers. Further, these two subject divisions had noticeably larger proportions of papers that use more than one dataset. In contrast, papers using only one dataset were the majority for the rest. In regards to data source and data type, datasets originated from government agencies as well as data generated from business operations accounted for the majority of the cited data, but significant differences existed among those subjects. As opposed to the frequently used business and series survey data, data generated from independent research projects as well as data that were non-continuous and one-time in nature were rarely reused in social sciences research. Based on the study findings, it is recommended that data services for social sciences research should focus more on large-scaled continuous data generated from governments and research institutions. A service mechanism that bridges users and data providers from public and private sectors would also enhance data reuse and increase the value of existing data.