Can career history be adequately drawn by self-reported survey?

采用邮件和网络调查的方式，对自述职业史调查数据中的技术性和非技术性问题进行了研究。不考虑数据结构的复杂性，从选民名册中随机选择的947个样本的实验表明，回复率令人满意，很少有项目无反应，没有显著的无反应偏差。然而，与那些给出预先编码的答案选项相比，被调查者在询问有关他们工作生涯的开放式问题时，显示出相对较高的项目无反应率。这些结果表明，如果通过邮件或网络进行职业生涯调查，则需要精心设计问卷和精心管理。

An experimental mail and internet survey was conducted to examine technical and non-technical problems of self-reported career history survey data. Regardless of the complexity of data structure, the experiment with 947 samples, randomly selected from electoral rolls, indicates a satisfactory response rate, scant item-nonresponses, and no significant nonresponse bias. However, respondents asked open-ended questions regarding their job career show relatively high item-nonresponse rate, comparing with those given pre-coded answer options. These results suggest that job-career survey need elaborate questionnaire and careful administration, if it would be conducted by mail or internet.