Objective :By analyzing the national physical surveillance data of Sichuan province in the year of 2000 ,2005 and 2010 ,this papertries to find out a more reasonable ,and represent‐ative sampling frame and the sensitive index of which effects the result .Methods :Taking the physical fitness level passing rate and some individual single indicator of the people that aged from20~59 and 60~69 as the objects of the study ,which originated from the national physi‐cal fitness surveillance in Sichuan province during the year of 2000 ,2005 and 2010 .Compared the results of different sampling frames .Results :There was difference between the results of 21 cities of Sichuan province ,especially in adults and the elderly .With the increasing of the physical surveillance data in the sampling frame ,the gap between the sampling frame and the 21 cities gradually reduced .And the number of indicators which were significant was gradually reduced .Conclusions and suggestions :In Sichuan province ,for example ,on the basis of 3 cities that currently existing ,it suggested that take 4 to 8 more cities into the national physical moni‐toring system .The sampling methods can learn from the theory of GIS spatial autocorrelation , establish the spatial weight according to geographic information or a property basis .%目的：对四川省2000年、2005年、2010年3次国民体质监测数据进行分析，比较不同抽样框之间数据的差异，尝试找出更为合理、更具有代表性的抽样框和影响抽样框结果的敏感指标，提高抽样效益和代表性。方法：以四川省3次（2000年、2005年和2010年）国民体质监测中20～59和60～69岁人群作为主要研究对象，以21市、州作为研究层面，选取体质达标率、代表性体质单指标等作为体质数据分析指标，将21个市、州分成不同的抽样框，对不同抽样框各项体质指标进行对比分析。结果：四川省国家监测点监测结果与四川省总体监测结果存在差异，这种差异主要体现在成年人和老年人；随着抽样框中监测城市的增多，抽样框与总体之间的差距逐渐缩小，且差异具显著性的指标数量逐渐减少。监测点12和监测点18与四川省总体之间差异提示，监测点12代表性与监测点18较为一致。结论与建议：以四川省为例，建议在目前现有的3个国家监测点基础上，考虑地域分布，再将4～8个市、州纳入国家国民体质监测体系，抽样方法可以借鉴GIS空间自相关理论，以地理信息或者某一属性为基础，建立空间权重，按照空间分层抽样策略设计流程进行抽样。
展开▼
