摘要
数据一致性是反映人口普查数据质量高低的重要指标。文章以第三、四、五、六次全国人口普查数据为基础,充分考虑人口死亡因素,比较同一出生队列人口在不同普查年份的人口规模,以此对人口普查数据一致性进行检验。研究中主要有三点发现:一是分时期进行分析,在"三普"至"六普"的28年间,九成以上出生队列人口存活率大于1,发生正偏离现象;二是对四次人口普查中的相邻两次普查依次进行比较,普查时处于0~90岁的91个出生队列人口偏离量逐渐增大,偏离率渐次上升;三是分性别来看,1982~2000年出生队列人口偏离程度存在显著性别差异。相当数量的出生队列人口出现系统性正偏离,表明在人口普查中存在较为严重的人口漏报、人口重报或死亡漏报。
Consistency is an important indicator of census data quality. This paper compares cen- sus population by birth cohort across the last four censuses taking into account age - specific death fac- tors. Results demonstrate that there is some inconsistency between data from the last four censuses. First, over 90% of the birth cohorts have positive deviations from 1982 to 2010, that is, cohort survival ratios are greater than 1. Second, deviations of birth cohorts 0 -90 between two adjacent censuses gradually increase. Third, there are significant gender differences on the cohort deviations over 1982 - 2000. There is a large number of birth cohorts having systematic positive deviations, indicating that there are serious problems of population underreporting, overreporting or death underreporting in the national population censuses.
出处
《人口研究》
CSSCI
北大核心
2013年第1期3-9,共7页
Population Research
关键词
人口普查
出生人口队列
偏离率
数据质量
Population Census, Birth Cohort, Deviation Rate,Data Quality