問卷調查引進中國大陸已有十餘年。採用這一西方社會科學研究報所進行的民意調查至少數以百計。然而，海內外對大陸民意調查的質量，如效度(validity)與信度(reliability)等，一直缺少研究。對許多大陸學者來說，此問題似乎是一技術性細節，不足以認真研究。就海外學者而言，這倒是挺重要的一個課題，但往往苦於無第一手資。本研究旨在通過剖析大陸問卷調查中的「單項不答」現象（包括「不知道」、「無意見」、「拒絕回答」等），而引發對此問題的重視與興趣。對十四項調查資料的統計分析表明，不答現象既受教育程度等個人因素的影響，也與問卷設計以及調查之等環境諸因素息息相關。 Why do some people say "I don't know," "I have no opinion," or simply refuse to say anything in a public opinion survey? The Western literature on public opinion research suggests two major causal factors for the non-response phenomenon: the respondent's individual characteristics (e.g., age, education, and occupation), and the survey's contextual characteristics (e.g., the length of the survey, the sensitivity of the question, and the setting of the interviews). The emphasis, however, has been on the individual factors. This study applies these theoretical predicators to a data pool of 14 surveys conducted in China during the 1980s. The results show that both individual and contextual factors are equally important in determining the occurrence of nonresponse. Based on the findings, we discuss the implications for analysis and interpretation of Chinese survey data.