抽樣調查研究方法的運用與檢討

APPLICATION OF SURVEY RESEARCH METHOD IN TAIWAN

本文先從抽樣調查研究的定義、特徵與缺點提出討論，並就調查研究在英國、美國和台灣的源起與發展進行扼要的說明。接著，作者針對台灣調查研究進行狀況提出說明與檢討。在抽樣方面，作者指出由於社會流動，以戶籍資料做為抽樣的依據造成偏差。在研擬問卷題目方面，在程序上國內研究者已經費時費心研擬推敲，但仍然有部份題目意義不明，信度和效度不佳，以致依變項被解釋量會偏低。在研擬問卷的問題上，作者進一步，就認知訪問和預試進行討論。在實地調查方面，作者強調訪員素質問題與訪員訓練。在資料分析方面，作者提出由於電腦計算的方便，就有不少無所用心或漫不經心的分析，失去學術意義。最後，作者特別呼籲國內要加強抽樣調查研究方法本身的研究，以提高國內調查研究的素質。

In the beginning of the paper, the author discusses definition, features and weakness of sampling survey research. Then, the author summarizes the origin and development of survey research in Britain, United States of America, and Taiwan. The major issues the author dealt with are sampling, questionnaire design, field work, and survey data analysis. In general, sampling frame based on household registration becomes troublesome due the biases from social mobility. Although many researchers try very hard to develop qualified questionnaire, some questionnaire items are still quite ambiguous or meaningless. Thus, the reliability and validity of some survey items are not high enough to assure powerful explanation of crucial variables. Cognitive interview and pilot study should be a part of questionnaire deign in order to improve the quality of survey items. More serious and well organized training program is needed for survey interviews in the field work of survey research. Some researchers tend to be careless or less thoughtful to do survey data analyses due to very convenient statistical analysis by using packages of statistical program and personal computer. Finally, the author suggests doing more researches on survey research method to improve quality of survey research in Taiwan.