Question

翻译下面一段文

Confounded Experiments in Advertising

A major value of a tightly controlled experiment is that the results are unambiguous: You can be confident that the independent variable led to changes in the dependent variable. Your conclusion about the effectiveness of the independent variable depends on minimal effects of extraneous factors – confounding variables. An advertising campaign presented an incorrect conclusion because the intended independent variable (Pepsi versus Coke) was confounded with another variable.

In the advertisement, people were asked to choose between the two cola drinks. In one series Pepsi was in a cup labeled S and Coke in a cup marked L. Most people chose the drink in S cup (Pepsi), so the conclusion was that cola drinkers prefer Pepsi. Is this a legitimate conclusion?

An experiment by Woolfolk, Castellan, and Brooks (1983) calls the advertising results into question. These authors reasoned that the labels on the cups may have had some influence on the choices made by the people in the advertisement. This supposition is based on the knowledge that labeling of various kinds can have a strong effect on consumer behavior. As a first step in their research, Woolfolk and her associates determined that college students like the letter S better than the letter L. Then the researchers had other students choose between the colas presented in cups labeled L and S. In this experiment, unlike in the advertisement, the letters were the only independent variable, because the type of cola in the cups was held constant. For half the subjects, both cups contained Pepsi; and for the other half, both cups contained Coke. Regardless of the type of cola in the cups, the students preferred cola S to cola L in 85% of the cases. Thus, it is possible that the conclusion in the advertisement was confounded by the preference for certain letters and did not result from the preference for different colas. Let the viewer beware!

Answer 1

广告中的困惑实验

一个严密控制实验的主要价值是，结果是明确的：你可以相信，独立变量导致了因变量的变化。你对自变量的有效性的结论取决于外在因素影响不大 - 混杂变量。一个广告呈现一个不正确的结论，因为预期的独立变量（百事可乐与可口可乐）是与另一个变量的混淆。

在广告中，人们被要求选择两家可乐饮料。在一个系列百事可乐杯是在一个标有S和一杯可乐标明属多数人选择了在S饮料杯（百事可乐），所以结论是，可乐饮用者更喜欢百事可乐。这是一个合法的结论呢？

一个由伍尔福克，城主，和布鲁克斯（1983）实验要求的广告效果产生疑问。这些作者的理由是，在杯子上的标签可能有一些在广告中人民作出的选择的影响。这个假设是基于知识，各类标签可能对消费者行为的强烈影响。作为其研究的第一步骤，伍尔福克和她的同事确定，像字母比英文字母L然后，研究人员选择了其他的学生在此实验中标记为L和学提出的可乐杯，不像在提高大学生广告，信件是唯一的独立变量，因为在可乐杯型为维持不变。半个世纪的主体，既包含百事可乐杯;以及另一半，两个杯子中的可乐。无论在杯子的可乐型，学生首选的可乐中85％的案件以可乐升。因此，它有可能在广告的结论是混淆了某些字母的偏爱，并没有产生于不同的可乐的偏好。让观众请注意