網路調查所面臨的問題與解決建議

Abstract

本論文的主要目的是透過基本研究方法的抽樣程序、樣本特質代表性、與外部效度等相關概念，探討網路作為問卷調查工具的疑慮與可行性。本文建議為避免過度擴大網路問卷調查的應用層面，危及調查結果的外部效度，研究的對象必須僅限於網路使用者，而非一般大眾；研究的議題也必須與網路使用具有相關性。同時，為解決網路樣本特質可能產生的誤差，必須對網路所蒐集的樣本進行適當的加權處理，包括人口統計變項與態度、生活形態變項。

The present paper investigates the questions about Internet-based surveys and attempts to explore the possible solutions by discussing some basic methodological concepts such as sampling, sample representativeness, and external validity. In order not to unduly extend the applicability of the Internet as a survey tool and consequently threaten the external validity, the present paper suggests that the target population for any Internet-based survey should be strictly limited to Internet users. The research topics for conducting any Internet-based survey should also be related only to Internet use. In addition, to reduce the possible bias caused by volunteer sample, certain post-survey weighting procedures must be performed, based on demographic as well as attitudinal and lifestyle variables.