因此，首先要迅速从短文中提炼出论点（the conclusion of the argument is that...), 论据（the author provides the evidence that/ cites the survey that/ employs the statistics that/) 论证（the author argues/reasons/that），这样一边审题，一边第一段写出来了。这一步很关键，关系到后面的反驳。
这就有了五宗罪了。特别是无理假设，最容易出现。 The argument is rested on the gratuitous assumption that…There is, however, no guarantee that this is the case, Nor does the author cites any evidence to support this assumption. In factor, more discreet inspections would reveal that many other factors far overweigh the factor the author heeds. For example, …
其次，论证有问题，论证最常见的就是说理论证和举例论证，说理论证的时候可能犯的错误是：
1。没有因果关系（no causal relationship) ，可能只是巧合，the author commits a fallacy of casual oversimplification. The fact A coincide with B does not lead to the conclusion that…without compelling evidence to support the casual relationship, the conclusion is not validated. 也可能根本没有关系，只不过是时间上先后顺序而已，the author commits a fallacy of assuming that just because A is followed by B, the latter event B causes the former event A. 也可能是因果关系不强烈，还有更强烈的因果关系，the only reason the author offers is the claim that…But it is entirely possible that the result can be attributed to other factors such as…Unless the author can rule out these and other factors relevant to …this argument commits a fallacy of causal oversimplification.
2。 不具备可比性（false analogy) 如果题目里面比较了两个情况，或者暗中把某个其他情况下的结果拿来用在新情况中，就会产生可比性的问题。The argument rests on the assumption that A is analogous to B. Although there are points of comparisons between A and B, differences clearly outweigh similarities, thus making the analogy highly invalid. For example, 说明有什么不同就可以了。
3。太绝对，否定中间可能性或者其他可能性( either-or choice) 有些题目的两个方面都有一定道理，两种情况可以不矛盾。The author assumes that A and B are mutually exclusive and there is no room for a middle ground. However, the author provides no reason that A and B cannot be reconciled. Common sense and observations tell us that combining A and B might generate even better results.
4。举例不充分：the evidence the author provide is insufficient to support the conclusion drawn from it. The statistics are collected from only a few samples. It may not be a good indicator of general case/trend. Unless the author can show that the samples selected are representative in general, he should not be so confident to conclude that…since it is also possible that…Hence, in face of such limited evidence, the conclusion is completely unwarranted
5.例子有问题：the poll cited by the author is too vague to be informative. The claim does not indicate who conducted the pool, who responded, or when, where, how the poll was conducted. Hence, lacking information about the percentage, the number, the composition, the education level, the age range, etc, it is unlikely to show that….
-----喜欢联结词，（in the first place, in sum, consequently, hence,) 这些可以让他审查文章的结构，句子的意思的递进，加深，转折，对比。看你文章挖掘的深度，广度，还有你的结构感。
----喜欢长句子，平均句子长度越长，得分越高。千万不要整篇文章都是简单句，主谓宾加上几个可怜的定语壮语，一句话就结束了。 长句子有几个窍门，第一个就是分句，尽量使用分句或者分词，使用减缩定语从句，前置壮语，（要注意逻辑主语一致），主语从句倒装结构，it is 形容词for someone that…第二个就是使用关联词，塑造平行结构，not only ,but also; either, or; 等等，尽量展开写成排比句。第三个就是用现在分词作伴随壮语，延长整个句子的意思，达到in other words的效果。 这样的句子经常出现在economist等杂志上面，一句话说完了，然后，thereby making, in effect leading to, thus resulting in ….第四个办法就是运用比较结构，unlike…in contrast to…第五个办法就是善于使用分号，分号隔开两个完整的简单句，如果可能，打造一两个有分号的句子，这样可以大大提高平均句子长度，但是要小心，分号不可乱用，不是任何两个句子，都可以把隔开他们的句号改成分号的，一定要有逻辑联系，比如，第一句话是generalization, 第二句话是具体的例子。或者前后是对比。
----喜欢例子，机器可以检查出来你有没有举例子的，不仅仅是检查“for example， such as ”, 如果你能在文章中写出 Microsoft, Hillary Clinton, 1998 Asian Financial Crisis, Big Bang Theory, Darwin’s evolution theory, 等等专有名词，特别是如果你的例子恰好就在这个题目标准答案的database中，分数能上一个档次（0。5）.注意，尽量用现实或者历史上的著名例子，特别是英美国家的人名，地名，大事的年份，不要怕拼写拿不准，机器能给你纠正。 次一点的例子就是自己组织的，注意，不是乱编的，是根据你的中心思想组织的例子，现实中可能存在的。 实在不行，就用both common sense and experience tell us that, as we all know, 最后没办法，可以放弃，但是三段中，至少有两个例子。
