15. The following memorandum is from the business manager of Happy Pancake House restaurants.

　　"Recently, butter has been replaced by margarine in Happy Pancake House restaurants throughout the southwestern United States. This change, however, has had little impact on our customers. In fact, only about 2 percent of customers have complained, indicating that an average of 98 people out of 100 are happy with the change. Furthermore, many servers have reported that a number of customers who ask for butter do not complain when they are given margarine instead. Clearly, either these customers do not distinguish butter from margarine or they use the term 'butter' to refer to either butter or margarine."Write a response in which you discuss what specific evidence is needed to evaluate the argument and explain how the evidence would weaken or strengthen the argument.

　　这个题目在13.10.1/13.12.4/14.3.9三次考试中都考到了这个题目，在题库的176个题目中考到3次的题目算是比较高频的题目了。那我们一起来分析下这个题目吧：

　　拿到题目，要找这个段落的结论是什么。应该是第二句：This change, however, has had little impact on our customers. 在这个段落中说到两个原因来证明结论：第一，only about 2 percent of customers have complained, indicating that an average of 98 people out of 100 are happy with the change. 第二，many servers have reported that a number of customers who ask for butter do not complain when they are given margarine instead. 其实这两个原因都不能充分地证明这个结论。

　　首先，我们可以先看看第一个原因，only about 2 percent of customers have complained,?indicating?that an average of 98 people out of 100 are happy with the change. 2%没有抱怨就一定代表100人中98人是乐于接受人造黄油吗?不一定，在100人中98人没有抱怨可能是因为他们自身的原因，觉得抱怨比较麻烦或者赶时间吃完就走了或者吃完不满意下次再也不来了，都有可能导致没有接受到所有的抱怨。

　　其次，第二个原因是，many?servers have?reported?that a number of customers who ask for butter do not complain when they are given margarine instead. 这里说“many”, 很显然，vague data, 一些人没有播报，但到底是多少呢?占服务生总量的百分比是多少?而且是不是就某个连锁店里的一些服务生这么说的呢?所以他们多少人、有没有代表性都是令人质疑的。并且后边还有一个“report”,就算前边的“many”很多人也很有代表性，但是他们不一定会“report”实情，因为他们是饭店的服务生，很有可能会隐藏有抱怨的事实，没说真话。

　　再次，either?these customers do not distinguish butter from margarine or they use the term 'butter' to refer to either butter or margarine."注意这里面出现的”either”、”or”字眼，它们是false dilemma非常标志性字眼，简称假两难。可能就算有抱怨也不是either、or里面的任何原因导致的，有可能是：他们是冲着这的就餐环境来的，上什么都无所谓;或者他们觉着这的dish price比较低;或者非常的tasty等等优点让他们不再计较是人造黄油还是天然黄油。

　　参考范文：

　　The above argument concludes that when Happy Pancake House replaced butter with margarine in their restaurants in the Southwest, there was no effect on their customers. To support this claim, they point the fact that only 2 percent of customers have complained about the change and that many servers say customers do not complain when given margarine instead of butter. They reason that customers generally fail to distinguish margarine from butter or use the term ‘butter’ to refer to both. However, their conclusions do not follow from the evidence presented.

　　The author claims that only 2 percent of customers complained, and that this indicates 98 out of 100 people are satisfied. First of all, the statistic itself is suspect. Where does the 2 percent come from? Perhaps they estimated their total number of monthly customers based on sales figures and then divided their number of monthly complaints by that number, but the reader can only guess. What if customers tend to only complain the second time an error is committed? Depending on how the statistic was calculated, this could mean that it only represents half the extent of customer dissatisfaction.

　　Second of all, even assuming the statistic is accurate, their assumption that 98 out of 100 people must then be happy does not logically follow. Some customers may not voice their dissatisfaction—perhaps the company’s protocol for filing complaints is too inconvenient for most customers to bother with, or perhaps dissatisfied customers stop dining at the restaurant altogether. Hence there are many possible scenarios in which complaint statistics do not accurately reflect customer dissatisfaction. The author could strengthen their argument by indicating how they arrived at the 2 percent statistic and why it is a valid measure of overall customer satisfaction.

　　The second piece of data used in the author’s argument is even worse. Dispensing with numbers altogether, they put forward a weakly worded claim that “many servers” reported “a number of customers” do not complain when given margarine after asking for butter. The errors here are essentially the same as those in the preceding discussion, just more grievous. To begin with, how many data points were used? What constitutes “many” servers? The sample size was likely too small to generate a valid statistic, so instead we are left with this nebulous phrasing. Perhaps only a handful of servers in a couple of restaurants reported this; we cannot know. Maybe they only looked for feedback in poorly performing restaurants where customer expectations were already low; again, we cannot know.

　　Furthermore, even if they were able to get a wide sample of restaurants and customers, it is possible that customers are reluctant to complain directly to servers because of the awkwardness it might cause. It also might be the case that butter and margarine are relatively unimportant to customers in comparison to the overall quality of their meal. We cannot assume their silence betokens a failure to distinguish butter from margarine, or that they must use the word ‘butter’ to refer to both spreads. Without some type of a control, this piece of data is essentially meaningless.

　　Customer complaints can be a useful form of feedback, but the author fails to use them reasonably in their argument. Eliminating butter from their restaurants may very well be an effective long-term business strategy for Happy Pancake House. However, the pieces of data provided are not persuasive enough to advise that course of action. (570 words)