| dc.contributor | 選研中心 | |
| dc.creator (作者) | 俞振華 | |
| dc.date (日期) | 2020-03 | |
| dc.date.accessioned | 1-七月-2024 15:30:43 (UTC+8) | - |
| dc.date.available | 1-七月-2024 15:30:43 (UTC+8) | - |
| dc.date.issued (上傳時間) | 1-七月-2024 15:30:43 (UTC+8) | - |
| dc.identifier.uri (URI) | https://nccur.lib.nccu.edu.tw/handle/140.119/152192 | - |
| dc.description.abstract (摘要) | 由於多數網路調查採用非機率樣本，因此其調查結果往往不被視為能用來推論母體特徵。近年來雖然西方調查學界針對網路調查的修正機制有諸多討論，但台灣的相關研究仍很缺乏。針對選舉與政治性議題調查，本研究旨在探討：透過結合機率樣本做為「對照組」（包括面對面調查或電話調查結果），各種不同修正機制在台灣的適用性。具體來說，本研究比較不同「入選機率調整法」（propensity score adjustment, PSA）應用時的良窳，並嚐試應用迴歸統計模型加上輔助變數的事後分層加權模式來修正網路調查結果。另外，本研究還將透過模擬，探討究竟使用多少的機率樣本做為「對照組」，即足夠修正網路調查結果。做為一項基礎研究，本研究希冀提出一些適用於台灣、並且有實用價值的網路調查修正機制，來解決非機率樣本的代表性問題。 | |
| dc.description.abstract (摘要) | As the majority of web surveys use non-probability samples, their results can hardly be used to make inferences about the population parameters. Over the past few years, although the western community of survey methodologists has lunched an extensive discussion regarding how to modify and improve web survey estimates, such discussion and relevant research is still meager in Taiwan. Using Taiwan’s survey data on electoral and political issues, this study tends to bridge the gap by comparing different modification approaches that use probability samples as a reference group (e.g., face-to-face or telephone surveys). Specifically, this analysis compares different types of propensity score adjustments (PSA) and develops a modeling approach that combines regression models with other auxiliary data for post-stratification as a way to modify web survey results. Additionally, this research tends to conduct a simulation analysis to explore the minimum sample size for the reference group dataset. In short, this study aims to develop multiple tools to reduce sample biases of web surveys and to identify some useful venues for the relevant future research in Taiwan. | |
| dc.format.extent | 116 bytes | - |
| dc.format.mimetype | text/html | - |
| dc.relation (關聯) | 科技部, MOST107-2410-H004-123, 107.08-108.07 | |
| dc.subject (關鍵詞) | 網路調查; 非機率樣本; 入選機率調整法; 事後分層加權 | |
| dc.subject (關鍵詞) | web survey; non-probability sample; propensity score adjustment; post-stratification | |
| dc.title (題名) | 網路調查結果修正機制之比較與模擬 | |
| dc.title (題名) | Options for Reducing Web Survey Bias | |
| dc.type (資料類型) | report | |
