中国农业普查事后质量抽查

一、ＰＥＳ数据分析模型由于抽查不只是对普查的重复 ,而是针对已进行的调查来收集“真实值”的一种方法。在中国 ,ＰＥＳ的入户访问是按与普查相同的环境来进行的 ,目的是要得到尽可能接近真实的调查结果。从ＰＥＳ得到的数据要与普查数据相比较。对两次调查的回答逐一比较 ,会得到如下的结果 :①普查和抽查回答一致 ;②两次调查的回答不一致 :回答的差别越大 ,普查估计的可靠性就越低。ＰＥＳ是为了检查和评估普查的数据质量。对ＰＥＳ回答误差的研究表明 ,回答误差也影响了ＰＥＳ数据。下面 ,我们将对两次调查同等对待。估计可靠性的指标…

The Afterward Qualitative Sample Survey of China Agriculture Census

The post-enumeration survey(PES) of the First National Agriculture Census in China(NAC) was carried out in order to check the quality of the census data at the national level. A sample of about 40 000 rural households was enumerated, census and PES records were matched one-to-one, and census-PES discrepancies on farm structure and management were computed. Almost 150 000 records of rural family members were matched, and individual response errors were computed. The most apparent errors on survey estimates were those of the interviewers. As far as respondents’ errors are concerned, is was determined that most of the census data are highly reliable, both at the farm and individual levels but some of the figures should not be published at the provincial or lower levels. Indications for future activities are given.