The purpose of a poll has nothing to do with whether one fraction is over represented vs. another. You need to work on your negation method and your reasoning skills. Focus on the conclusion of an argument.
The purpose of a poll has nothing to do whether one fraction is over represented vs. another. You need to work on your negation method and your reasoning skills. Focus on the conclusion of an argument.
-- by 会员 sdcar2010 (2012/3/17 12:53:11)
嗯，先用SDCAR教的方法试着做下。。。 The U.S. census is not perfect: thousands of Americans probably go uncounted(background). However, the basic statistical portrait of the nation painted by the census is accurate. (Conclusion)Certainly some of the poor go uncounted, particularly the homeless（Concession）; but some of the rich go uncounted as well, because they are often abroad or traveling between one residence and another.（Premise）
我说下我的思路：如果否定B（即不可能合理的把所有的美国人分成穷人和富人，那么明显有一部分人会被分为中间阶层，至少是未分类）。那题干的conclusion：the basic statistical portrait of the nation painted by the census is accurate.就不成立。你提到过“The purpose of a poll has nothing to do whether one fraction is over represented vs. another.”但是这个题是Census（人口普查），不是poll（民意测验，投票选举），前者更客观，后者更主观（呃，如果我对这两个词的理解不到位还请见谅哈）。如果fraction在体现the basic statistical portrait中不重要的话，那什么重要呢？Anyway，我还是想举个例子rich40%，Middle20%（或者说未分类，以否定B为前提，则一定会有此类），poor40%，差错率为10%，则rich36%，middle20%，poor36%。。。一个很明显的有差错的basic statistical portrait就是这个社会的基尼系数会大大降低。可能扯远了，总之我认为一个集合的统计特征的偏差绝对不可能通过一两个简单的假设就修正过来的。原因很简单，统计特征这个概念太大了！
You were wrong again because you added one MORE assumption than the answer choice and the original passage allow: There is NO mistake when counting the middle! Think about it. Further, you did not understand the meaning of "uncounted", which is linked to both the total number of the counted population and that of each fraction. As long as each fraction has the same statitistical error, the total representation of the population is correctly represented by the data.
Census and poll are all statistical analysis. For this question which is discussing the number and percentage, census and poll are the same.
"You were wrong again because you added one MORE assumption than the answer choice and the original passage allow: There is NO mistake when counting the middle! "这个假设确实是我为了证明原题目有问题而加上去的。既是所谓的为了证明结论（这个结论是“否定Ｂ题目中的结论照样成立”，而不是题干中的结论）错误，而从大类中找出一小类，证明这一小类不正确，那么基于大类的结论自然不攻而破。所以加上这个假设是可行的。
但是显然，这个“As long as each fraction has the same statitistical error”的假设是题目外的。正如你说的，只要这个假设不成立（题干中没有这个假设），那么这个结论"the total representation of the population is correctly represented by the data”就有问题。（因为这是个充分必要的关系，所以我直接取反可以推出正确的结论）
Hehe. What you said above proves EXACTLY that B) is not NECESSARY since your added assumption after negating B) ONLY covers one of several possible scenarios.
B) is a sufficient assumption. But B) is not necessary.
京公网安备11010202008513号 京ICP证101109号 京ICP备12012021号
楼主
发表于 2012-3-17 20:52:01
