文章詳目資料

中國統計學報 EconLit

  • 加入收藏
  • 下載文章
篇名 大型面訪調查的抽樣架構變革:從戶籍抽樣到地址抽樣
卷期 56:4
並列篇名 An Exploratory Study of the Transition from Register- Based Sampling to Address-Based Sampling in Large Face-to-Face Surveys
作者 謝淑惠詹大千劉雪涵傳仰止
頁次 075-097
關鍵字 地址抽樣戶籍抽樣個人資料保護法最小統計區等距抽樣address-based samplingregister-based samplingPersonal Information Protection Actbasic statistical areainterval samplingEconLitTSCI
出刊日期 201812

中文摘要

台灣於2012年10月1日開始施行「個人資料保護法」(以下簡稱個資法),政府機關爲了 符合個資法規定,對於個人資訊的維護與管理更加嚴格,導致申請戶籍資料用於面訪調查的難 度大爲提高,社會調查以戶籍資料作爲抽樣淸冊的壓力越來越大。爲了謀求長遠的解決方式, 有些大型的面訪調查已於2015年開始採用地址抽樣(Address-based Sampling, ABS)的方 式代替過去30年來慣用的戶籍抽樣。本研究將探討地址資料的涵蓋率和以最小統計區(Basic Statistical Area, BSA)的抽樣設計。爲了評估可能的影響,我們以北北基宜地區爲主,模擬等 距抽樣、區域中心點等距抽樣、隨機中心點等距抽樣等不同方案,每個方案各抽出648個門牌地 址,並作爲進一步分析衡量不同方案下的樣本離散程度。以祈能對全面地址抽樣的未來盡一份 心力,此外仍有許多關於地址抽樣的研究課題亟待進一步探討。

英文摘要

Ever since the Personal Information Protection Act (PIP Act) became effective in Taiwan in October 2012, large social surveys have faced increasing pressure to avoid over reliance on the population register as the sampling frame. To comply with the PIP Act, the Ministry of the Interior (MOI) has started imposing strict restrictions over taking personal information from the population register as the base for sampling survey respondents. For face-to-face surveys, in particular, the use of personal information is becoming unreliable or even infeasible. As a pioneer in seeking alternative sampling schemes, the Taiwan Social Change Survey (TSCS) began to develop new address-based sampling (ABS) strategies after having relied on governmental register-based sampling (RBS) for over three decades. As the flagship social survey in Taiwan and one of the oldest and largest general social survey series in the world, the TSCS needs to resolve some of the emerging issues caused by the new sampling strategies. We discuss some of these issues, such as the coverage of a doorplate database and use of a small geographic unit of basic statistical areas in the sample design. To examine the possible effects, we simulate samples in the northern part of Taiwan by taking three approaches, respectively: systematic sampling; interval sampling, in which the starting point is a regional center; and interval sampling, where the starting point is a random center. We further measure the trend for each approach by analyzing a sample of 648 doorplates. Based on our findings at this critical early stage of the transition from RBS to ABS, we urge researchers to address additional relevant evolving issues of using the ABS in Taiwan.

相關文獻