- UID
- 170718
- 热情
- 16407
- 人气
- 21388
- 主题
- 435
- 帖子
- 24365
- 精华
- 35
- 积分
- 31550
- 分享
- 0
- 记录
- 0
- 相册
- 2
- 好友
- 4
- 日志
- 0
- 在线时间
- 18142 小时
- 注册时间
- 2009-1-30
- 阅读权限
- 30
- 最后登录
- 2023-9-23
    
升级   57.75% - UID
- 170718
- 热情
- 16407
- 人气
- 21388
- 主题
- 435
- 帖子
- 24365
- 精华
- 35
- 积分
- 31550
- 阅读权限
- 30
- 注册时间
- 2009-1-30
|
not419 发表于 2015-7-13 14:47 
在这个统计里,主要用到了华人的姓氏比较集中这个特点。
有点意思,值得研究 ...
这个姓氏的统计基本站的住脚(样本够大的时候)。
Mr Twyford said Labour had worked out the buyers' likely ethnicity by comparing their surnames against electoral roll data, which covers ethnicity and other information for 1.4 million Aucklanders.
The method could not identify any one person as Chinese but was accurate across large groups. For instance, only 40 per cent of people called Lee in the city were Chinese, compared to almost 96 per cent of people called Li.
也就是说,在上次统计中,比如有X%的某个姓氏说自己是chinese,那么在这个房屋买卖统计中出现的这个姓氏,也可以认为有X%属于chinese,然后加以计算。 这个在统计分析上没有大的错误。 |
|