Abstract: Age structure is the most fundamental population structure, thus, assessing the accuracy of age reporting is an important prerequisite for census data usage. Using the data from population censuses and sample surveys, this paper employs the methods of comparing with other data sources, cohort survival ratio, modified Whipple index, and Sign Test to investigate the accuracy of age data in the 7th population census. The number of annual births estimated from population aged 0-9 in the 7th census and that releaseal by the National Bureau of Statistics are quite close, which preliminarily suggests that there is no obvious underreporting in the lowage group of the 7th census. The cohort survival ratios abnormally fluctuate in the age group of 10-40, reflecting the underreporting in the lowage group, the overreporting in youth group, and the larger overreporting of women of childbearing age than males in the 6th census. In addition, this study adopts the modified Whipple index and Sign Test method to examine the accuracy of age reporting in the 7th census. The results show that the 7th census age data is of high quality without obvious age heaping at the national level.

摘要： 年龄结构是最基础的人口结构，对年龄数据准确性进行评估是使用普查数据的重要前提。利用历次人口普查和抽样调查数据，通过与其他来源数据比较、队列存活率法、修正惠普尔指数、符号分布法等考察第七次全国人口普查年龄数据质量。将普查0~9岁人口回推得到2011~2020年出生人数，和国家统计局公布的年度出生人数比较，二者很接近，尤其是2015年后几乎一致，推断第七次全国人口普查低龄组不存在明显漏报。对10岁及以上人口队列存活率的观察发现，10~40岁队列存活率存在异常波动，主要是由第六次全国人口普查低龄组漏报、青年组重报和育龄女性重报大于男性等造成。修正惠普尔指数和符号分布法的分析结果均表明，第七次全国人口普查在全国层面上不存在特定的年龄尾数指向，年龄申报质量很高。

关键词: 队列存活率, 修正惠普尔指数, 符号分布法, 第七次全国人口普查