作者: George Duncan
关键词:
摘要: Government statisticians are intrigued by the possibility of accessing administrative data as a way to enhance survey and census data. Survey rich in attributes, but they exhibit sampling errors. More consequentially, demonstrate non-sampling errors, often because nonresponse, also responses that incomplete or inaccurate, frequently due respondent’s inability recall. Additionally, panel subject attrition. Administrative can help compensate for these problems importantly data, such social security files, include almost everybody. Further, supplementing statistical be cost-effective, an enormous amount already collected support functional operations government agencies, so potentially available. In many instances, respondent burden surveys lessened certain attribute values may available from records. Some U.S. federal e.g., Social Security Administration (McNabb et al., 2009) have linked with broaden its demographic socioeconomic measures improve quality See National Research Council (2005), pp.45–48 additional examples discussion benefits linkage.